A Study to Explore the Causes of Higher Notification of Tuberculosis in Adult Females in the Province of Khyber Pakhtunkhwa (KPK), Pakistan

A Study to Explore the Causes of Higher Notification of

Tuberculosis in Adult Females in the Province of Khyber

Pakhtunkhwa, Pakistan

A thesis submitted to The University of Manchester for the degree of

Doctor of Philosophy

in the Faculty of Biology, Medicine and Health

2016

Muhammad Aziz

School of Health Sciences

LIST OF CONTENTS

LIST OF CONTENTS ...... 2

LIST OF TABLES ...... 12

LIST OF FIGURES ...... 16

LIST OF ABBREVIATIONS ...... 18

ABSTRACT ...... 20

DECLARATION ...... 21

ACKNOWLEDGEMENT ...... 23

DEDICATION ...... 25

THE AUTHOR ...... 27

LIST OF PUBLICATIONS AND CONFERENCE PROCEEDINGS ...... 30

CHAPTER 1: INTRODUCTION ...... 31

1.1 Introduction ...... 31

1.2 Burden of Tuberculosis ...... 32

1.2.1 Global Burden ...... 32

1.2.2 Regional Burden...... 33

1.2.3 National Burden of TB in Pakistan ...... 34

1.2.4 Tuberculosis in Khyber Pakhtunkhwa ...... 35

1.3 Summary ...... 37

CHAPTER 2: HEALTH SERVICES IN PAKISTAN ...... 38

2.1 Introduction ...... 38

2.2 Demography of Pakistan ...... 38

2.3 Geography of Pakistan ...... 40

2.4 Geography of Khyber Pakhtunkhwa ...... 41

2.5 Healthcare in Pakistan ...... 44

2.5.1 Primary Healthcare...... 45

2.5.1.1 Basic Health Unit ...... 45

2.5.1.2 Rural Health Centre ...... 46

2.5.2 Secondary Healthcare...... 46

2.5.2.1 Tehsil Headquarter Hospital ...... 46

2.5.2.2 District Headquarter Hospital ...... 47

2.5.3 Tertiary Healthcare...... 47

2.6 Public Health Services ...... 48

2.7 Private Health Services ...... 49

2.8 The Expanded Programme on Immunisation (EPI) in Pakistan ...... 50

2.9 National Tuberculosis Control Programme ...... 51

2.10 Summary ...... 53

CHAPTER 3: LITERATURE REVIEW ...... 54

3.1 Introduction ...... 54

3.2 Aim of the Literature Review ...... 54

3.3 Methodology of Literature Review ...... 55

3.3.1 Systematic Review ...... 55

3.3.2 Search Strategy...... 55

3.3.2.1 Search Strategy for Objective 1 ...... 56

3.3.2.2 Search Strategy for Objective 2 ...... 56

3.3.2.3 Search Strategy for Objective 3 ...... 57

3.3.2.4 Search Strategy for Objective 4 ...... 57

3.4 Inclusion Criteria ...... 59

3.5 Exclusion Criteria ...... 59

3.6 Summary of Literature Selection ...... 60

CHAPTER 4: FINDINGS OF THE SYSTEMATIC REVIEW “TUBERCULOSIS

IN PAKISTAN” ...... 72

4.1 Epidemiology of TB in Pakistan ...... 72

4.1.1 Introduction ...... 72

4.1.2 Age and Gender...... 74

4.1.3 Educational Status ...... 76

4.1.4 Socio-Economic Status ...... 77

4.1.5 Ethnic Origin ...... 78

4.1.6 Summary ...... 78

4.2 Risk Factors for Disease Acquisition ...... 79

4.2.1 Introduction ...... 79

4.2.2 Disease Risk Factors ...... 79

4.2.2.1 Household Contact ...... 79

4.2.2.2 Overcrowding ...... 80

4.2.2.3 Smoking ...... 81

4.2.2.4 Malnutrition ...... 82

4.2.3 Cultural Risk Factors ...... 82

4.2.3.1 Male Dominant Culture ...... 82

4.2.3.2 Summary ...... 83

4.2.4 Knowledge Based Risk Factors ...... 83

4.2.4.1 Patient's Knowledge about TB ...... 83

4.2.4.2 Physicians’ Knowledge about the disease ...... 87

4.3 Summary of the Literature Review ...... 93

CHAPTER 5: METHODOLOGY ...... 94

5.1 Introduction ...... 94

5.2 Aim of the Study ...... 94

5.3 Objectives of the Study ...... 94

5.4 Research Questions ...... 94

5.5 Hypothesis ...... 95

5.6 Methodology for Objective One ...... 95

5.6.1 Study Setting and Target Population...... 95 5

5.6.2 Study Design ...... 96

5.6.3 Data Collection...... 96

5.6.4 Data Quality ...... 97

5.6.5 Analysis ...... 97

5.6.6 Ethics ...... 97

5.7 Methodology for Objectives Two and Three (Study Design) ...... 98

5.7.1 Study Setting and Target Population...... 98

5.7.2 Study Design ...... 98

5.7.3 Sample Size ...... 99

5.7.3.1 Inclusion Criteria ...... 99

5.7.3.2 Exclusion Criteria ...... 100

5.7.4 Date Collection Tools ...... 100

5.7.4.1 Baseline Data ...... 101

5.7.4.2 Demographic Data ...... 101

5.7.4.3 Risk Factors ...... 101

5.7.4.4 Knowledge of TB...... 101

5.7.5 Validity of Data Collection Tool...... 102

5.7.5.1 Content Validity ...... 102

5.7.5.2 Concurrent Validity ...... 102

5.7.5.3 Piloting the Questionnaire ...... 102

5.7.6 Data Collection...... 103 6

5.7.6.1 Face to face interview ...... 103

5.7.6.2 Double Data Entry Strategy ...... 104

5.7.6.3 Analysis ...... 104

5.7.7 Ethical Approval ...... 104

5.7.7.1 Ethics Committees ...... 104

5.7.7.2 Informed Consent ...... 105

5.7.7.3 Confidentiality ...... 105

5.8 Methodology for Objective Four ...... 105

5.8.1 Research Plan ...... 105

5.8.2 Study Setting and Target Population: ...... 106

5.8.3 Study Design ...... 106

5.8.4 Sample Size for Objective Four ...... 106

5.8.5 Date Collection Tools ...... 106

5.8.5.1 Demographic Data...... 107

5.8.5.2 Knowledge of TB...... 107

5.8.6 Data Collection...... 107

5.8.7 Dual Data Entry...... 107

5.8.8 Analysis ...... 107

5.8.9 Ethics ...... 107

5.9 Summary ...... 108

CHAPTER 6: RESULTS - DESCRIPTIVE EPIDEMIOLOGY ...... 109

6.1 Introduction ...... 109

6.2 TB Patients Registration from 2002 to 2010 ...... 109

6.3 Annual Incidence of TB in KPK ...... 112

6.4 Distribution of TB Patients by Age and Gender ...... 115

6.5 Classification of TB in Patients ...... 115

6.6 Notification of New Sputum Positive Cases by Age Structure ...... 117

6.7 Summary ...... 121

CHAPTER 7: RESULTS - RISK FACTORS ...... 122

7.1 Introduction ...... 122

7.2 Distribution of Patients by Districts ...... 123

7.3 Age Distribution of Patients ...... 124

7.4 Educational Status of Notified Cases in Higher versus Lower Notification

Districts ...... 125

7.5 Distribution of Patients by Family Size in Higher versus Lower Notification

Districts ...... 126

7.6 Distribution of Patients by Income ...... 126

7.7 Distribution of Notified Cases by Age of the Youngest Child ...... 128

7.8 Distribution of Notified Cases by Type of Fuel Used for Cooking ...... 128

7.9 Distribution of Patients by House Type ...... 129

7.10 Distribution of Notified Cases by Amount of Time Spent Outside the House. ...

...... 129

7.11 Summary Table of Gender Differences and Risk Factors for TB (Model-1). 131

7.12 Summary Table of Risk Factors for TB Notification by High and Low

Notification Districts. ( Model-2) ...... 133

7.13 Summary Table of Multiple Logistic Regression ...... 134

7.14 Summary ...... 137

CHAPTER 8: RESULTS - PATIENTS’ KNOWLEDGE ...... 138

8.1 Introduction ...... 138

8.2 Gender Distribution of Patients by Higher and Lower Notification Districts 138

8.3 Distribution of Patients by Urban and Rural Residence ...... 139

8.4 Knowledge of Patients about the Causes of TB ...... 140

8.5 Knowledge of Patients about the Transmission of TB ...... 141

8.6 Knowledge of Patients about the Spread of TB ...... 142

8.7 Knowledge of Patients about the Signs and Symptoms of TB ...... 143

8.8 Knowledge of Patients about Drug Resistance of TB ...... 144

8.9 Overall Knowledge of TB Patients...... 145

8.9.1 Patients’ Knowledge about the Cause, Symptoms, Transmission and

Spread of TB ...... 146

8.9.2 Patients’ Knowledge about the Treatment and Complications of TB..... 146

8.10 Summary ...... 150

CHAPTER 9: RESULTS - PHYSICIANS’ KNOWLEDGE ...... 151

9.1 Introduction ...... 151

9.2 Demographic Distribution of Physicians ...... 151

9.3 Qualification and Service of Physicians ...... 152

9.4 Knowledge about the Causes, Spread and Diagnosis of TB ...... 152

9.5 Knowledge about Treatment of TB Patients ...... 153

9.6 Knowledge about Follow-up and Complications of TB ...... 153

9.7 Overall Knowledge of Physicians ...... 154

9.8 Summary ...... 157

CHAPTER 10: DISCUSSION ...... 158

10.1 Introduction ...... 158

10.2 Discussions of Methodology and Limitations ...... 160

10.3 Discussions of Descriptive Epidemiology (2002-2010) ...... 164

10.4 Discussions of Results Risk Factors ...... 166

10.4.1 Sample of Patients ...... 166

10.4.2 Age Distribution of Patients ...... 166

10.4.3 Educational Status of Patients ...... 167

10.4.4 Socio-economic Status of Patients ...... 169

10.4.5 Malnourishment and Social Empowerment of Women ...... 170 10

10.4.6 Analysis of Logistics Regressions ...... 171

10.5 Patients’ Knowledge of TB ...... 173

10.6 Physicians’ Knowledge of TB ...... 176

CHAPTER 11: IMPLICATIONS, CONCLUSIONS AND RECOMMENDATIONS ..

...... 178

CHAPTER 12: FUTURE RESEARCH ...... 181

CHAPTER 13: REFERENCES ...... 182

CHAPTER 14: APPENDIX ...... 194

14.1 Ethical Approval (The University of Manchester) ...... 194

14.2 National TB Control Programme Ethical Approval Letter ...... 196

14.3 Training of Medical Technicians for Data Collection Project ...... 200

14.4 Data Collection Instruments ...... 205

14.5 Conference Presentations ...... 216

14.6 University Positions for Students ...... 222

14.7 National TB Control Programme Data Collection and Reporting Instruments ...

...... 226

14.8 PRISMA Checklist ...... 235

Words Count: 48,656

LIST OF TABLES

Table 1: Five Higher TB Occurrence States in 2009 (15) ...... 33

Table 2: Comparison of Gender Differences of TB (NSS+ve) Case Notification of KPK with other Provinces in Pakistan from 2002-2010(6) ...... 36

Table 3:Percentages of Gender Differences of TB (NSS+ve) Case Notification of KPK with other Provinces in Pakistan from 2002-2010 (6) ...... 37

Table 4: Pakistan's National Age Structure (25) ...... 39

Table 5: Sex Ratio of Pakistan (25) ...... 39

Table 6: Pakistan Demographic Summary (26) ...... 39

Table 7: Distribution of KPK Population According to Rural /Urban Ratio, Sex Ratio and Area by District in 1998 Census (29) ...... 42

Table 8: Provincial Comparison of Health Services in Pakistan (38) ...... 48

Table 9: A Comparison of the Public and Private Sector Health Care System in Pakistan

(39) ...... 50

Table 10: Pakistan EPI Childhood Immunisation Plan (42) ...... 51

Table 11: Search Strategy and Terms ...... 58

Table 12: Findings from the Different Search Strategies...... 61

Table 13: Summary Table of the Literature describing Gender Differences of TB in

Pakistan ...... 63

Table 14: Summary Table of the Literature of Risk Factors for Tuberculosis ...... 68

Table 15: Summary Table of the Literature on Patient’s Knowledge about Tuberculosis

...... 70

Table 16: Summary Table of the Literature on Physician’s Knowledge about TB ...... 71

Table 17: Five Higher and Five Lower Notification Districts of KPK ...... 98

Table 18: Annual Case Notification of TB by Districts in KPK, 2002-2010 ...... 111

Table 19: Incidence of TB from 2002 to 2010 in KPK, Pakistan ...... 113

Table 20: Distribution of TB Patients by Gender and Type of Disease, 2002-2010 .... 116

Table 21: Notification of Smear Sputum Positive Cases by Age Structure in KPK, 2002-

2009 ...... 117

Table 22: Notification of New Cases by Gender and District, 2002-2010 ...... 119

Table 23: Distribution of Patients by Districts ...... 123

Table 24: Distribution of Patients by Age...... 124

Table 25: Educational Status of Patients by Higher against Lower Notification Districts

...... 125

Table 26: Distribution of Notified Cases by Family Size in Higher and Lower

Notification Districts ...... 126

Table 27: Income Distribution of Notified Cases by Gender ...... 127

Male versus Female; Chi Square = 23.77, p=<0.001Table 28: Income Distribution of

Notified Cases in Higher and Lower Notification Districts ...... 127

Table 29: Distribution of Notified Cases by Age of the Youngest Child ...... 128

Table 30: Distribution of Notified Cases of Fuel Used for Cooking...... 129 13

Table 31: Distribution of Notified Cases by House Type ...... 129

Table 32: Gender Distribution of Notified Cases by Amount of Time Spent Outside the

Home ...... 130

Table 33: Distribution of Notified Cases by Amount of Time Spent Outside the Home between High and Low Notification Districts ...... 130

Table 34: Gender Difference and Risk Factors for TB (Model-1)...... 132

Table 35: Differences in Risk Factors for TB Notified Cases between Low and High

Notification Districts ...... 133

Table 36: Summary Table of Multiple Logistic Regressions ...... 136

Table 37: Gender Distribution of Patients in Higher and Lower Notification Districts138

Table 38: Distribution of Patients by Urban and Rural Residence in Higher and Lower

Notification Districts ...... 139

Table 39: Knowledge of Patients about the Causes of TB by Gender in High and Low

Notification Districts ...... 140

Table 40: Knowledge of Patients about the Transmission of TB by Gender in High and

Low Notification Districts ...... 141

Table 41: Knowledge of Patients about the Spread of TB by Gender in High and Low

Notification Districts ...... 142

Table 42: Knowledge of Patients about Signs and Symptoms of TB by Gender in High and Low Notification Districts ...... 143

Table 43: Knowledge of Patients about Drug Resistance of TB by Gender in High and

Low Notification Districts ...... 144 14

Table 44: Overall Knowledge of TB Patients by Gender in High and Low Notification

Districts ...... 145

Table 45: Summary of Responses of the Participants Regarding Causes, Symptoms and

Spread of TB by Gender in High and Low Notification Districts ...... 147

Table 46: Summary of Responses of the Patient's Knowledge about Treatment and

Complications of TB by Gender in High and Low Notification Districts ...... 148

Table 47: Age Distribution of Physicians in Higher and Lower Notification Districts 151

Table 48: Distribution of Physicians by Qualification ...... 152

Table 49: Distribution of Physicians by Service and District of Notification ...... 152

Table 50: Summary of Questions Asked for Assessing Physicians’ Knowledge about the

Diagnosis, Treatment and Complications of TB...... 154

Table 51: Knowledge of Physicians in Relation to Causes, Management and

Complications and Summary of Overall Knowledge...... 155

Table 52: Summary of Overall Knowledge of Physicians by Age, Years of Service,

Experience, Training and Qualifications...... 156

LIST OF FIGURES

Figure 1: Global Distribution of TB (New TB Cases) in 2007 (15)...... 33

Figure 2: TB Burden in the Eastern Mediterranean Region (18) ...... 34

Figure 3: Age Structure of Pakistan's Population (24) ...... 38

Figure 4: Map of Pakistan (25) ...... 40

Figure 5: Map of Khyber Pakhtunkhwa (30) ...... 43

Figure 6: Queen Elizabeth II visit to Swat in 1961(31)...... 44

Figure 7: PRISMA Flow Diagram ...... 61

Figure 8: Annual Case Notification from 2002 to 2010 ...... 111

Figure 9: Total Case Notification by Districts in KPK, from 2002 to 2010 ...... 112

Figure 10: Cure and Success Rates of TB in KPK, 2002-2010 ...... 113

Figure 11: Total Number of Annual Deaths from TB in KPK, 2002-2010 ...... 114

Figure 12: Annual Default Rate of TB Patients in KPK, 2002-2010...... 114

Figure 13: Distribution of TB Patients by Gender, 2002-2010 ...... 116

Figure 14: Distribution of TB Patients by Disease Type, 2002-2010 ...... 116

Figure 15: Distribution of New Smear Positive Cases by Age Structure ...... 118

Figure 16: Population Pyramid of New Sputum Positive Cases, 2002-2010 ...... 118

Figure 17: Distribution of TB Patients by Gender and District, 2002-2010 ...... 120

Figure 18: Distribution of Patients Notified by Higher and Lower Districts ...... 121

Figure 19: Gender Differences in TB Notification in Higher and Lower District ...... 124

Figure 20: Correct Responses of Female Patients about the Causes, Symptoms and

Spread of TB ...... 149

Figure 21: Correct Responses of Female Patients about the Treatment, and

Complications of TB ...... 149

LIST OF ABBREVIATIONS

AFG Afghanistan

JID Journal of Infectious Diseases

AJPH American Journal of Public Health

BHU Basic Health Unit

BTB Bovine tuberculosis

CDR Case detection rate

DHO District Health Officer

DHQ District Headquarter Hospital

DOTS Direct Observed Therapy Short Course

DPT Diphtheria, pertussis and tetanus

EMR Eastern Mediterranean Region

EPI Expanded Programme on Immunisations

ERS Electronic Reporting System

ESR Erythrocytes sedimentation rate

EXTB Extra Pulmonary Tuberculosis

GAVI Global Alliance for Vaccines and Immunisation

GPs General Practitioners

HBCs High Burden Countries

HND Higher Notification Districts

KPK Khyber Pakhtunkhwa

LHW Lady Health Worker

LND Lower Notification Districts

MBBS Bachelor of Medicine Bachelor of Surgery

MCH Mother and Child Health

MD Doctor of Medicine

MDG Millennium Development Goals

MDR Multi drug resistant

TB Mycobacterium tuberculosis

ND Notification Districts

NGO Non-governmental Organisation

NIMS Northern Institute of Medical Sciences

NTCP National Tuberculosis Control Programme

NWFP North West Frontier Province

PRISMA-P Preferred Reporting Systematic reviews and Meta-analysis Protocol

PKR Pak Rupees

PPs Private Practitioners

PTB Pulmonary Tuberculosis

RHC Rural Health Centre

TB Tuberculosis

TSR Treatment Success Rate

UNHCR United Nations High Commission for Refugees

WHO World Health Organisation

The University of Manchester Muhammad Aziz PhD: November 2015 ABSTRACT A Study to Explore the Causes of Higher Notification of Tuberculosis in Adult Females in the Province of Khyber Pakhtunkhwa (KPK), Pakistan. Background: Tuberculosis (TB) continues to be major cause of death for adult females in Pakistan, which ranks as the fifth highest country in the world with a TB burden. Globally, TB occurrence, morbidity and mortality is higher in males. However, the notification of TB in Khyber Pakhtunkhwa (KPK) and Balochistan is higher in females as compared to males. This study aims to explore the risk factors associated with the higher notification of TB in women in KPK province of Pakistan. Aim : To explore the reasons behind the variation of Tuberculosis Notification between males and females in the province of Khyber Pakhtunkhwa (KPK), Pakistan. Objectives: 1. To describe the epidemiological characteristics of confirmed TB cases in KPK by age, gender and geographical location between 2002 and 2010. 2. To determine the risk factors associated with TB in the province of KPK. 3. To assess the knowledge of new diagnosed TB patients. 4. To evaluate the level of knowledge, attitude and practice of physicians working in TB diagnostic centres (TDCs) in KPK. Methods: 1. We retrospectively collected data for all registered TB patients between 2002 and 2010 using data from the National TB Control Programme (NTCP) in KPK, Pakistan. We analysed the data to show the distribution of TB by age, gender, type of disease, geographical location and treatment outcome. 2. We carried out a cross-sectional study from 1st July 2012 to 30th September 2012, identifying newly diagnosed TB patients from ten districts (five high notification and five low notification districts) in PKP to determine the risk factors associated with the diagnosis of TB and knowledge assessment of TB patients in higher versus lower notification districts. 3. We carried out a cross sectional study of a sample of general practitioners in these districts to assess their knowledge and management of TB cases. Results: Through a descriptive analysis of nationally collected data, we confirmed that female patients in KPK were at a higher risk of contracting TB compared to males. We determined that the majority of female patients in higher notification districts were illiterate, unemployed, poor and living in households of low socioeconomic status when compared to women from low notification districts. A strong association was noted for the interaction between education and gender (OR = 0.16), suggesting that more educated women were around 6 times less likely to be in a higher notification area (1/0.16) compared to a low notification area. It was also observed from the logistic regression model that income level (OR = 0.42), anaemia (OR = 0.45) and unemployment (Or = 0.23) were also associated with being in a high notification district in the province of KPK. Lower notification districts with female gender were used as reference category for the model. Finally, knowledge of general practitioners about the treatment, follow up and complications of TB was poor. Discussion: Our results suggest that differences in socio-economic factors are implicated in the differential notification of TB between men and women and between high and low notification districts. Poor knowledge of physicians in the treatment and management of TB may also be a contributory factor. This has implications for the future management of TB control programs in these areas and in Pakistan more generally. 20

DECLARATION

I undertake that no portion of the work referred to in the thesis has been submitted in support of an application for another degree or qualification of this or any other university or other institute of learning.

Muhammad Aziz

The following four notes on copyright and the ownership of intellectual property rights must be included as written below:

I. The author of this thesis (including any appendices and/or schedules to this thesis) owns certain copyright or related rights in it (the “Copyright”) and s/he has given The University of Manchester certain rights to use such Copyright, including for administrative purposes.

II. Copies of this thesis, either in full or in extracts and whether in hard or electronic copy, may be made only in accordance with the Copyright, Designs and Patents Act 1988 (as amended) and regulations issued under it or, where appropriate, in accordance with licensing agreements which the University has from time to time. This page must form part of any such copies made.

III. The ownership of certain Copyright, patents, designs, trademarks and other intellectual property (the “Intellectual Property”) and any reproductions of copyright works in the thesis, for example graphs and tables (“Reproductions”), which may be described in this thesis, may not be owned by the author and may be owned by third parties. Such Intellectual Property and Reproductions cannot and must not be made available for use without the prior written permission of the owner(s) of the relevant Intellectual Property and/or Reproductions.

IV. Further information on the conditions under which disclosure, publication and commercialisation of this thesis, the Copyright and any Intellectual Property and/or Reproductions described in it may take place is available in the University IP Policy (see http://documents.manchester.ac.uk/DocuInfo.aspx?DocID=24420), in any relevant Thesis restriction declarations deposited in the University Library, The University Library’s regulations (see http://www.library.manchester.ac.uk/about/regulations/) and in The University’s policy on Presentation of Theses

ACKNOWLEDGEMENT

I would like to thank everyone who has helped and supported me during my

PhD study in Manchester. First of all, I would like to thank to the committee members who were more than generous with their expertise and precious time. I am particularly grateful for their support when both of my supervisors left the University in second year of my PhD. The committee identified my current supervisor with both a public health and clinical background, which was very important for my study. I would like to express my deep appreciation to my supervisor Professor Aneez Esmail, who accepted me in the middle of my PhD. His critical review, inputs and constructive comments moved me forward and enabled me to complete my PhD thesis. I will never forget your kind support during the hard and tragic time I passed through from August to December

2013. I am also thankful to my co supervisors Professor Mark Woodhead and Dr David

Baxter for their valuable inputs and advice. I would also like to thank Professor Sara

O’Brien who was my main supervisor at the start of my PhD and for her assistance in creating a strong foundation for my study. I would like to thank my postgraduate administrators Christine Burns, Tasleem Hanif and Amina Saeed for helping and supporting me throughout my study in the University of Manchester.

I would like to thank Dr Catherine Reed, Programme Director of Postgraduate

Courses in Public Health and Mr Stewart Taylor, MPH e-learning Technologist,

University of Manchester for their support and cooperation and giving me access to online MPH courses of Epidemiology, Biostatistics, Evidence Based Practice,

Qualitative Research methods and Health System in Low Income Countries. These courses helped me a lot in completion of my research project.

I am also thankful to David Springate, Statistical Research Assistant, Primary

Care Research Group, Institute of Population Health, University of Manchester, for helping me in the development of the logistic regression model. 23

I would like to thank Dr Malek Shah, Manager, Scholarship Division and

Brother Shaharuqul Haq, Scholarship Officer at the Islamic Development Bank

Scholarship Programme for financing my PhD and who supported me during my PhD.

I am also thankful to Dr Abdul Latif and Dr Ubaid Hussain, Project Directors, and Dr Maqsood Ali, Project Manager, TB Control programme Khyber Pakhtunkhwa for their cooperation and support. I am also thankful to all the Executive District

Officers Health (EDOH) and District TB Control Officers for their continuous and outstanding support during the field work. I am also thankful to all the Medical

Technicians, Lady Health Workers and Area Supervisors for their efforts towards the successful completion of the data collection project.

On a personal note, I would like to thank to all those TB patients who participated in this study and who made this study possible.

I must also acknowledge all my friends, teachers, trainers and librarians who assisted, advised and helped me throughout my PhD. I am also grateful for the generous support of my friend Dr Salik Javeed for proof reading, feedback and constructive comments.

Finally, this was an enormous PhD research project with extensive financial and technical requirements. Therefore, I am sincerely thankful to the financial support of the Islamic Development Bank, Jeddah, Saudi Arabia for the award of the PhD scholarship for my research. I would also like to thank The University of Manchester for financing parts of this PhD and the Provincial TB Control Programme KPK for providing technical support, which helped me particularly in the successful completion of the data collection phase

DEDICATION

First of all, I am grateful to dedicate this thesis to my parents, without their support, encouragement and constant love I could not have completed this PhD.

Secondly, I would like to dedicate this thesis to my teachers Mr Muhammad

Neem, Mr Abdul Salm, Mr Gul Zaman, Prof Mehfoz Jan Abid, Prof Zia-ur-Rehman and

Abdul Raouf, who were involved in my previous education and I acknowledge that their enthusiasm and support is one of the reasons why I am here today and submitting my thesis for PhD.

Finally, I dedicate this work and special thanks to my wife and children who suffered a lot during my studies, particularly when I was away for field study. I believe that without your cooperation, engagement and patience, I could not have completed this milestone.

CERTIFICATE OF APPROVAL

This is to certify that the work presented in this thesis "A Study to Explore the

Causes of Higher Notification of Tuberculosis in Adult Females in the Province of

Khyber Pakhtunkhwa, Pakistan" submitted by Dr Muhammad Aziz was carried out by the candidate under my supervision. The information obtained from other sources has been duly acknowledged in the thesis.

University of Manchester Prof Aneez Esmail November 2015 Supervisor

THE AUTHOR

I have completed a two years Master's degree in Public Health (MPH) in 2009 from Abasyn University, Peshawar, Pakistan, charted by the Government of Pakistan and recognised by the Higher Education Commission (HEC) of Pakistan. I graduated from Ayub Medical College Abbottabad, Pakistan in 2006, in medicine (MBBS), which is recognized by the HEC and Pakistan Medical & Dental Council (PMDC) Islamabad.

Before starting my PhD I was working as a medical doctor at the United Nations High Commission for Refugees (UNHCR) and was an assistant professor in the Department of Community Health Sciences & Research at Northern Institute of Medical Sciences, (MINS College of Medicine), an institution that is involved in the areas of medical education and research.

I am interested in the research aspects of community health sciences (Public Health) and health care delivery management systems. This interest developed during my MBBS and MPH endeavours under the supervision of Prof Zia-ur-Rehman and Dr Irfan who were based in Ayub Teaching Hospital Abbottabad and Abasyn University Peshawar. My MPH dissertation was titled ‘EPIDEMIOLOGY OF DOG BITE CASES ATTENDING THE ACCIDENT & EMERGENCY DEPARTMENT OF AYUB TEACHING HOSPITAL, ABBOTTABAD’ which sought to study the epidemiological characteristics of dog bite injuries, the knowledge, attitude and practices of the subjects regarding dog bites. This work contributed towards providing reliable data to the national institute of health (NIH) in Pakistan regarding dog bite injuries.

In 2007, on annual tuberculosis day, I participated in a training workshop organised by the National TB Control Programme (NTCP). During the presentation, Dr Ghafor, the programme manager of NTCP showed us a slide describing the differential notification of males (42%) and females (58%) with respect to TB. He further explained that the notification of Tuberculosis is higher in males globally and nationally. However, in Khyber Pakhtunkhwa, there was a higher female notification rate compared to males. He posed the question, "Can I have anyone amongst you to investigate scientifically the causes of this higher notification". This sparked my interest and I started collecting information about the problem. In 2008, I completed the preliminary background work and submitted a research proposal on the issue to the Islamic Development Bank (IDB) for 3 Years PhD Merit Scholarship. I was given the award in 2009. After securing the scholarship, I applied to study for a PhD to several

UK universities and received offer letters from the University of Warwick, University of Leicester, Imperial College London and the University of Manchester. Before, accepting the offers, I visited all the above universities and chose the University of Manchester.

University Positions

Peer Mentor for the School of Community Based Medicine 2011-12 Postgraduate Student Representative (PGR) 2011-12 (See Appendix 14.6 on page 222)

My Training & Development in the University of Manchester Event Unit Code Title Event Venue Hours Date Graduate Teaching STDU, 2nd Floor, FMHSS2401 Assistants/Demonstrator 13/03/2012 Humanities Bridgeford 7 Training Street Building Research Ethics Room 3.204, University FMHSS1202 Application: University & 29/11/2011 4 Place NHS Introduction to Statistics room 4.2, Roscoe FMHSS1302 15/11/2011 12 using SPSS Building Room 4.38, Simon FMHSS1021 Mini First Year Workshop 03/11/2011 4 Building Hanson Room, Getting the most out of FMHSS2105 14/10/2011 Humanities Bridgeford 4 research conferences Street Building Academic Writing in 2.03, Mansfield Cooper FMHSS1100 12/10/2011 27 English Building Communicating your Room 4.38, Simon FMHSS2101 06/10/2011 3 Research with the Public Building Health System in Low MPH e-Learning MEDN62212 2011-12 150 Income Countries Technology Qualitative Research MPH e-Learning MEDN63121 2011-12 150 Methods Technology Fundamentals of MPH e-Learning MEDN60991 2011-12 150 Epidemiology Technology MPH e-Learning MEDN60982 Biostatistics 2011-12 150 Technology MPH e-Learning MEDN60041 Evidence Based Practice 2011-12 150 Technology Communicate with G306A/B, Jean FMHSS1106 09/09/2011 3 Confidence McFarlane Building Introduction to Effective Room 4.50, Simon FMHSS1102 08/02/2011 6 Presentation Skills Building Critical analysis of research Room 1.010, Roscoe FMHSS1301 03/02/2011 4 papers Building Systematic Review Room 4.2, Roscoe FMHSS2303 12/01/2011 7 Training Building Effective Publications: Room 5.206, University FMHSS2300 Taking the sting out of peer 10/12/2010 4 Place review Academic Writing in Room 3A, Simon FMHSS1100 13/10/2010 30 English Building An Introduction to Room 2.825, Stopford FMHSS Statistical Modelling with 28/09/2010 48 Building Stata

First Year: Introduction to Day 1 & 2: Room 1.010, FMHSS1010 Research Speed PhD and 23/09/2010 10 Roscoe Building MD Course

LIST OF PUBLICATIONS AND CONFERENCE PROCEEDINGS

44th World Conference on Lung Health, Organised by the International Union against Tuberculosis and Lung Diseases (The Union), Paris France 30 October-03

November 2013 (Poster Presentation) Title: Are women at higher risk of TB? An epidemiological study of prevalence and risk factors for TB in Khyber Pakhtunkhwa,

Pakistan (Appendix 14.5 on page 216).

45th World Conference on Lung Health, Organised by the International Union against Tuberculosis and Lung Diseases (The Union), Barcelona Spain 28 October-01

November 2013 (Poster Presentation) Title: Gender and tuberculosis: A prospective study exploring risk factors for higher notification of tuberculosis (TB) in adult women in Khyber Pakhtunkhwa (Appendix 14.5 on page 216).

First Islamic Development Bank (IDB) Scholar's Scientific Conference,

University of Cambridge UK 22 May 2014 (Oral and Poster Presentation) Title: Did you know, women are at higher risk of Tuberculosis in Khyber Pakhtunkhwa (KPK),

Pakistan? An epidemiological study of prevalence and risk factors for TB in KPK,

Pakistan (Appendix 14.5 on page 216).

6th International Conference on Health Care and Life Sciences Research

(ICHLSR), Imperial College London, 18-19 September 2015 (Oral and Poster

Presentation) Title: Tuberculosis in Adult Women: A systematic Review of Gender

Differences in Tuberculosis Notification in the province of Khyber Pakhtunkhwa,

Pakistan (Appendix 14.5 on page 216).

Faculty PGR Showcase: Faculty of Medical and Human Sciences, School of

Medicine, University of Manchester 09 April 2014 (Poster Presentation) Title: Are women at higher risk of TB? An epidemiological study of prevalence and risk factors for TB in Khyber Pakhtunkhwa, Pakistan (Appendix 14.5 on page 216).

CHAPTER 1: INTRODUCTION

1.1 Introduction

Tuberculosis (TB) continues to be a major cause of morbidity in adult females in Pakistan, which globally ranks as the fifth highest country with a TB burden (1-3).

Tuberculosis was affirmed as a major worldwide health problem in 1993 by the World

Health Organisation (WHO) because of concerns about the alarming development of multi-drug resistant tuberculosis. The WHO has developed a series of initiatives which includes an extensive programme of Direct Observed Therapy Short Course (DOTS).

The DOTS programme has now become the recommended treatment approach not only for the detection and cure of TB, but also as a means of global tuberculosis control.

However, despite implementation of the DOTS programme across Pakistan, the disease remains uncontrollable. One of the criticisms of the WHO’s approach is that it is not gender aware and there is emerging evidence of the differential impact of risk of infection, progression of disease, case fatality and access to treatment which is mediated by gender difference. For example, according to TB data reported to the National

Tuberculosis control programme (NTP) in Pakistan, the notification of TB has been higher in females when compared to men in the province of Khyber Pakhtunkhwa

(KPK), Pakistan since 2000 (4). In the status of TB drug resistance, Abu-Amero has argued that this has also led to a negligence of the gender aspects of TB in relation to research and control of the disease (5).

Globally, TB case notification is higher in males as compared to females but in

KPK female patients are at a higher risk of TB than male patients (6).This study, therefore, aims to explore the risk factors associated with higher notification of TB in women in KPK province of Pakistan with a view to identifying areas where supplementary action could be provided to strengthen the planning, administration and

31 implementation of the National TB Control Programme (NTP). In addition, this study will explore the knowledge and practice of general practitioners (GPs) in the province of Khyber Pakhtunkhwa as it relates to the diagnosis and treatment of TB. This is important because GPs are the main group of healthcare professionals who are responsible for the diagnosis and management of TB in many parts of Pakistan. The purpose of this study will be to understand and explain gender differences in the notification and treatment of TB.

1.2 Burden of Tuberculosis

Tuberculosis (TB) was declared a global emergency in 1993 by WHO (7).

Efforts were subsequently made to expand partnerships and bring all stakeholders on board in order to promote effective control of the disease, and recognise correlation of socioeconomic conditions with new cases of TB (8-10).

Some of the highest areas of TB incidence are areas of Africa, Asia and Latin

America where the gross national product (GNP) is low (3)(11). An estimated eight million people are infected with TB each year, 95% of those living in developing countries. Globally it is estimated that nearly 2 million people die each year from TB and its associated complications (12).

1.2.1 Global Burden

In 2012, the WHO estimated that there were 8.6 million cases of TB globally.

This is equivalent to 122 cases per 100,000 of population. In the same year, 1.3 million people died because of this curable disease (13).

An expert group, commissioned by the WHO, has prepared estimates of the number of cases which were categorized by age and gender as part of an update to the

Global Burden of Disease study (14). This showed that the 35% of all cases of TB were attributed to women, giving an estimated number of 3.3 million cases worldwide.

Figure 1: Global Distribution of TB (New TB Cases) in 2007 (15).

In terms of distribution, the largest numbers of TB occurrences in 2012 were in

Asia (58%) and Africa (27%). The Americas had the lowest reported incidents, accounting for 3% of all cases (13). The 22 high burden countries (HBCs), which included Pakistan, accounted for 81% of all estimated cases worldwide. China and India together, accounted for 26% of all cases, with India contributing to 21% of cases and representing one-fifth of all TB cases globally (16)[Table 1].

Table 1: Five Higher TB Occurrence States in 2009 (15) State Cases Population in 1000 Incidence Rate India 2.4 million 1,236,687 176 China 1.1 million 1,377,065 73 South Africa 0.60 million 52,386 1000 Indonesia 0.50 million 246,864 185 Pakistan 0.50 million 179,160 231

1.2.2 Regional Burden

Nine out of twenty countries of the WHO Eastern Mediterranean Region

(EMR), contributed to total of 94% of TB cases in the region. Pakistan has the highest 33 prevalence of patients (43%), while Yemen has the lowest (3%). Figure 2 shows TB burden by country in the EMR. However Gillini and Seita in an article about the control and epidemiology of TB in the EMR region suggested that TB data for Pakistan and

Afghanistan had probably been underestimated because of the low detection rate in the two countries (17).

Figure 2: TB Burden in the Eastern Mediterranean Region (18) Iran Others 6% 6% Egypt Yemen 4% 3% Morocco 5% Pakistan 43% Somalia 6% Iraq 7% Sudan Afghanistan 8% 12%

1.2.3 National Burden of TB in Pakistan

In terms of TB incidence, Pakistan was ranked number one among the high burden countries (HBCs) in the EMR in 2013 and ranked fifth worldwide. According to the WHO, Pakistan’s estimated tuberculosis incidence for all forms of the disease is

330/100,000 and for sputum positive TB cases the rate is 80/100,000 population per year (1)(12). Despite this high incidence, TB remains one of the neglected areas for health investment. Figures show that Pakistan’s expenditure on health as percentage of

GDP (2.2%) is significantly lower than neighbour countries such as Afghanistan

(7.6%), Iran (5.7%) and India (4.2%) (19). In 2014, Pakistan spent 114 million US$ on

TB of which 30% was donated by international organisations. This lack of investment may be one of the reasons why the prevalence of TB in Pakistan contributes nearly 44% of the total tuberculosis burden in the Eastern Mediterranean Region. Data from the

WHO also shows that TB is responsible for 5.1% of the total national disease burden in 34

Pakistan with a significant impact on socio-economic status of the population (20). TB treatment in Pakistan is free but indirect costs such as travel to health facilities and loss of income may cause socio-economic problems for low earning families. TB commonly occurs in the most productive age group of 15 to 45 years, considered to be the most economically active (19). Although DOTS was introduced in Pakistan in 1995, it was only in 2000 that the Ministry of Health began its wider implementation. It was probably as a result of the national implementation of DOTS through the NTP that the detection rate of TB and DOTS coverage doubled (21). However, despite this increase in DOTS population coverage, from 24% in 2001 to nearly 63% in 2003, the positive case detection rate has remained low with only 17% in 2003. This suggests that many patients do not have access to DOTS even within the designated DOTS areas because the TB burden remains high (22).

1.2.4 Tuberculosis in Khyber Pakhtunkhwa

More than half (22,000) of the estimated 40,000 TB patients who registered with 214 public investigative and 1,029 management centres in 24 districts across the

Khyber Pakhtunkhwa province in 2009 were females. Two-hundred fifty eight (n = 258) of these females were reported to have died of TB in the Khyber Pakhtunkhwa and

Federally Administered Tribal Areas (FATA) (23). Over a nine year period, the notification of new sputum positive (NSS +ve) cases remained higher in females of

KPK comparing to data reported from Punjab, Sindh and Pakistan, where the notification was higher in males. Similarly, figures from Balochistan province also showed higher notification of females as compared to males from 2002 to 2010

(6)[Table 3][Table 3].

Table 2: Comparison of Gender Differences of TB (NSS+ve) Case Notification of KPK with other Provinces in Pakistan from 2002-2010(6) Pakistan KPK Punjab Sindh Balochistan Year Male Female Total Male Female Total Male Female Total Male Female Total Male Female Total 2002 6521 6725 13246 1239 1720 2959 1011 917 1928 3029 2372 5401 347 512 859 2003 9820 9504 19324 1907 2492 4399 1985 1720 3705 4634 3431 8065 391 585 976 2004 15113 14520 29633 2465 3422 5887 3962 3459 7421 6851 5224 12075 783 1076 1859 2005 23661 23086 46747 3473 5081 8554 9039 7986 17025 8915 7140 16055 1168 1552 2720 2006 33461 31126 64587 4278 5776 10054 16188 13998 30186 10646 8166 18812 1367 1976 3343 2007 45123 42442 87565 5029 6838 11867 25405 22523 47928 12006 9582 21588 1564 2094 3658 2008 50587 48341 98928 5308 7209 12517 30374 27558 57932 12253 10030 22283 1519 2040 3559 2009 51974 49354 101328 5547 7572 13119 31199 27963 59162 12517 10193 22710 1368 1908 3276 2010 53033 51444 104477 6191 7945 14136 31478 28644 60122 12775 11274 24049 1474 2147 3621

Table 3:Percentages of Gender Differences of TB (NSS+ve) Case Notification of KPK with other Provinces in Pakistan from 2002-2010 (6) Pakistan Punjab Sindh Balochistan KPK Year M(%) F (%) M (%) F (%) M (%) F (%) M (%) F (%) M (%) F (%)

2002 49.23 50.77 52.44 47.56 56.08 43.92 40.40 59.60 41.87 58.13

2003 50.82 49.18 53.58 46.42 57.46 42.54 40.06 59.94 43.35 56.65

2004 51.00 49.00 53.39 46.61 56.74 43.26 42.12 57.88 41.87 58.13

2005 50.62 49.38 53.09 46.91 55.53 44.47 42.94 57.06 40.60 59.40

2006 51.81 48.19 53.63 46.37 56.59 43.41 40.89 59.11 42.55 57.45

2007 51.53 48.47 53.01 46.99 55.61 44.39 42.76 57.24 42.38 57.62

2008 51.14 48.86 52.43 47.57 54.99 45.01 42.68 57.32 42.41 57.59

2009 51.29 48.71 52.73 47.27 55.12 44.88 41.76 58.24 42.28 57.72

2010 50.76 49.24 52.36 47.64 53.12 46.88 40.71 59.29 43.80 56.20

1.3 Summary

This chapter has briefly examined TB at global, regional, national and local levels. It shows that the TB burden in Pakistan is the greatest in the EMR. Data from

KPK province shows that this burden seems to fall disproportionally on women. In part, local disease trends reflect health service utilization, which will be discussed in the next chapter.

CHAPTER 2: HEALTH SERVICES IN PAKISTAN

2.1 Introduction

The goal of this section is to highlight the health services and infrastructure in

Pakistan to give a better understanding of the study area. The geography of Pakistan and

Khyber Pakhtunkhwa state will also be described.

2.2 Demography of Pakistan

According to WHO 2010 report the estimated population of Pakistan is about

180 million, making it the world`s sixth most populated country with a 1.95% population growth rate. The literacy rate of females (36.0%) is significantly lower than males (63.0%) (23).

Figure 3: Age Structure of Pakistan's Population (24)

Figure 3 shows the age and gender structure of Pakistan's population. Horizontal axis shows gender distribution of the population and vertical axis represents the age groups of the population.

Table 4: Pakistan's National Age Structure (25) Age Group Total (%) Rural (%) Urban (%)

0-14 43.40 45.06 40.09 15-44 42.14 40.15 46.15 45-64 10.93 10.98 10.83 > 65 3.53 3.81 2.93 Table 4shows the distribution of urban and rural population of Pakistan by different age groups.

Table 5: Sex Ratio of Pakistan (25) Sex Ratio per 100 Females Age Group Total (%) Rural (%) Urban (%) 0-14 108.0 102.9 110.6 15-44 92.8 90.6 96.6 45-64 97.0 93.2 106.1 > 65 127.8 136.4 113.3 Overall 101.6 100.4 103.9

Table 6: Pakistan Demographic Summary (26) Description Figures Population 1,822,143,000 estimated 2013 Growth rate 2.03 Birth rate (estimated births/1000) 27 Death rate (estimated deaths/1000) 7.86 Deaths due to tuberculosis (per100,000 69 population) Literacy total (Male, Female) 49.90 ( 63.00, 36.00) Unemployment 7.4 Life Expectancy 65.6 Gross national income per capita (PPP int $) $4,920

Table 6 shows the WHO latest Pakistan's demographic data about population, literacy, unemployment and gross national income.

2.3 Geography of Pakistan

Pakistan, located in the north-western part of the South Asia, became an independent state as a result of the division of British India on August 14, 1947. After the Indo-Pakistani War of 1947-48, Pakistan annexed Azad (Free) Kashmir. The northeast part of the subcontinent where the majority of the population were Muslims also became part of Pakistan as East Pakistan. Hence, East and West Pakistan were separated by 1,600 km of Indian Territory. In December 1971, East Pakistan became the independent state of Bangladesh (25).

Figure 4: Map of Pakistan (25)

Pakistan has four provinces Punjab, Sindh, Khyber Pakhtunkhwa and

Baluchistan. The total land area of Pakistan which occupies a position of great geostrategic importance is estimated about 803,940 square kilometres (310,401 sq mi).

It occupies a great geostrategic importance as it borders with Afghanistan (2,430 km) in the west, China (523 km) in the northeast, India (2,912 Km) on the east and the Arabian

Sea (1,064 km) on the south. The estimated population of Pakistan is about

1,822,143,000 with 2.7% growth rate. This makes it in the top ten of the world population growth rate (27).

2.4 Geography of Khyber Pakhtunkhwa

Khyber Pakhtunkhwa (KPK) is the third largest populated province of

Pakistan. The total land area of KPK is about 74,251 square kilometres (28,773 sq mi).

Peshawar, Abbottabad, Mardan, Bannu, Dera Ismail Khan, Dir, Swat and Chitral are the main districts and the main cities are Bannu, Kohat, Haripur, Mansehra, Peshawar,

Mardan, and Abbottabad (28).

Table 7: Distribution of KPK Population According to Rural /Urban Ratio, Sex Ratio and Area by District in 1998 Census (29) Urban Rural Sex Growth Area Population Propor Propor Rate Ratio tion tion (1981-98) District Males per % per sq.kms % 000s % % % Hundred Annum Females NWFP 74521 100 17736 100 16.88 83.12 105 2.82 Abbottabad 1967 2.64 881 4.97 17.93 82.07 100 1.82 Bannu 1227 1.65 676 3.81 7.04 92.96 107 2.81 Battagram 1301 1.75 307 1.73 - 100.00 107 -0.58 Buner 1865 2.50 506 2.85 - 100.00 100 3.86 Charsadda 996 1.34 1022 5.76 18.86 81.14 108 2.88 Chitral 14850 19.93 319 1.80 9.61 90.39 103 2.52 D.I.Khan 7326 9.83 853 4.81 14.75 85.25 111 3.26 Hangu 1097 1.47 315 1.78 20.42 79.58 96 3.25 Haripur 1725 2.31 692 3.90 11.95 88.05 100 2.19 Karak 3372 4.52 431 2.43 6.47 93.53 96 3.26 Kohat 2545 3.42 563 3.17 27.00 73.00 101 3.25 Kohistan 7492 10.05 473 2.67 - 100.00 124 0.09 Lakki 3164 4.25 490 2.76 9.57 90.43 104 3.16 Lower Dir 1583 2.12 718 4.05 6.18 93.82 98 3.42 Malakand 952 1.28 452 2.55 9.55 90.45 107 3.36 Mansehra 4579 6.14 1153 6.50 5.32 94.68 98 2.40 Mardan 1632 2.19 1460 8.23 20.21 79.79 107 3.01 Nowshera 1748 2.35 874 4.93 25.96 74.04 109 2.90 Peshawar 1257 1.69 2019 11.38 48.68 51.32 111 3.56 Shangla 1586 2.13 435 2.45 - 100.00 106 3.27 Swabi 1543 2.07 1027 5.79 17.45 82.55 101 2.96 Swat 5337 7.16 1258 7.09 13.83 86.17 106 3.37 Tank 1679 2.25 238 1.34 15.00 85.00 109 3.13 Upper Dir 3699 4.96 576 3.25 3.98 96.02 103 2.76 Table 7 shows the distribution of rural and urban populations of KPK by districts, where Battagram, Buner, Kohistan, and Shangla districts are 100 % rural residence.

Key to text colour in table:

 = Districts out of Research Area

 = Districts with Lower Female TB Notification

 = Districts with Higher Female TB Notification 42

Figure 5: Map of Khyber Pakhtunkhwa (30)

The region differs in its natural features. The southern part is mostly dry and rocky while the northern part mostly consists of woodlands and green plains. The typical weather in this region varies from hot summers to freezing winters. In spite of these large variations in weather, agriculture is both viable and it plays an important role in the regional economy.

Kalam, Upper Dir, Naran and Kaghan are all known for their magnificent mountainous landscape and tourism remains an important component of the local economy. Swat was called "the Switzerland of the east" by Queen Elizabeth II during her visit to the Yusafzai state of Swat in 1961 (31).

Figure 6: Queen Elizabeth II visit to Swat in 1961(31).

Figure 6 shows Queen Elizabeth II with Price Philip, Miangul Jehanzeb Khan (Wali-e- Swat) and Qudratullah Shahab (Urdu writer and bureaucrat)

According to the 1998 census report, the population of KPK is 17,734,645 (of whom 52% are males and 48% are females). There are approximately 1.5 million

Afghan refugees living in the province. Two-thirds of the population are classified as

Pashtuns who make the largest ethnic group in the area (32).

2.5 Healthcare in Pakistan

The development of the health services in Pakistan can be traced to policy developments implemented at the time of partition. The Health Survey and

Development Committee, popularly known as the Bhore Committee, reported in 1946 that the future state of Pakistan should assume full responsibility for all preventive measures and treatment necessary for the health protection of the population (33).

Recent policy developments have resulted in attempts to provide a range of comprehensive services by the State and attention has also been paid to the distribution

44 of medical benefits and the ability to pay for health care irrespective of a person's economic situation. Politically, the state ascribes to the notion that health care is a fundamental right which should be available to all irrespective of ability to pay (25).

The health care delivery system in Pakistan can be classified into three categories: primary, secondary and tertiary care.

2.5.1 Primary Healthcare

Basic health unit (BHU) and rural health centre (RHC) are primary healthcare facilities mostly providing preventive care for immunisation such as for polio and TB.

Family planning and maternal and child health services are also available in primary healthcare facilities.

2.5.1.1 Basic Health Unit

The basic health unit (BHU) was created in 1980 and established by the government during 1985-86 in every union council. Lady health workers (LHWs) linked to BHUs reinforce the health system at community level, where BHUs are the main upward referral centres for rural and district health hospitals. The government also decided to provide clinics and medical supplies in all larger union councils in 1991-92.

Staff at BHUs consists of a medical doctor, medical technician, health technician, vaccinator, midwife and LWH. BHUs provide health services for 5000 to

10000 of the population. BHUs have morning out-patients department (OPD) from

Monday to Friday between 0800 to 1500 hours (33-34). The total annual budget of the

BHU is around 1.33 million Pakistani Rupees (USD 133,000; 1 US$= 100 Pak Rupees at the time of writing) (35).

2.5.1.2 Rural Health Centre

There was a significant urban rural divide with fewer government-supported services in the rural areas. The country’s second five-year plan (1960-65) required the establishment of 150 rural health centres (RHC) over a period of five years in West

Pakistan alone. By December 2000, this number had increased to 572 RHCs (35).

In the same way that BHUs, are organised, RHCs also provide preventive services. RHCs not only provide health services to a population of 25000 to 50000 but are also focal points for community and health authorities for meeting and training of health personnel. The staffing of RHCs consist of senior medical officer, medical officer, women’s medical officer, medical technicians, midwifes and LHWs (34)(36).

2.5.2 Secondary Healthcare

Tehsil headquarter hospitals (THQ) and district headquarter hospital (DHQ) are secondary healthcare facilities, providing primary curative health services for outpatients and inpatients.

2.5.2.1 Tehsil Headquarter Hospital

These are located in the centre of Tehsils providing preventive, promotive and primary curative services for 100,000 to 150,000 residents. THQ are the first level referral facilities for RHCs and BHUs. They are equipped with morning OPDs as well as 24-hour emergency and ambulance services. THQ also have 40 to 60 bedded facilities for inpatients divided to children, surgical and medical wards. All THQs have operative and diagnostic facilities that include medical laboratory and radiology equipment. Personnel in THQ consist of medical superintendent (MS), specialist doctors, medical officers, nurses, medical technicians, midwives and other supportive staff (34).

2.5.2.2 District Headquarter Hospital

DHQ hospitals are located in every district, providing curative care to half to one million residents. DHQs are secondary level referral facilities for THQs, RHCs and

BHUs. They have both morning and evening OPD with 24-hour ambulance services.

DHQs are125 to 250-beded facilities divided into children’s, surgical, medical, orthopaedic, cardiology and maternity wards. DHQ hospitals also have dental, intensive care units and operating theatres (OT). Personnel in DHQ consist of MS, specialist doctors, medical officers, dental surgeon, pharmacist, medical technicians, OT technicians, nurses, Medico legal officer (MLO), midwives and other supporting staff

(36).

2.5.3 Tertiary Healthcare

Tertiary healthcare hospitals (THC) are mainly located at the centre of provinces or big cities, providing specialised consultative care with more advanced diagnostic facilities. Patients usually referred to THC from DHQs, THQs, RHCs and

BHUs. Teaching hospitals are an example of highly skilled tertiary care level treatment hospitals in Pakistan (37). Teaching hospitals are mainly curative with medical education, research and training facilities. THCs have bedding capacity ranging from

600 to 1200, with specialised obstetrics and gynaecology, cardiology, surgical, medical, paediatric, skin, orthopaedics and 24-hour emergency units. THCs also have advanced diagnostic and waste management services. Tertiary care hospitals provide advanced treatment services such as plastic surgery, angiography, angioplasty, coronary artery bypass etc. Both morning and evening outpatient clinics are available for outpatients and inpatients in THCs. Personnel in THC consists of MS, deputy medical superintendents, casualty medical officers (CMOs), specialist doctors, medical officers,

47 trainee medical officer, house officers, medical technicians, OT technicians, nurses, managers, medical directors and other supporting staff.

Table 8 shows that Punjab and Khyber Pakhtunkhwa provinces have better health services than Sindh and Baluchistan (38).

Table 8: Provincial Comparison of Health Services in Pakistan (38) Health Provinces Total Facility PUNJAB SINDH KPK Baluchistan DHQs 28 11 15 18 72 THQs 57 44 10 0 111 Dispensaries 1,006 309 623 652 2590 TB Centres 46 1 24 9 80 MCH 404 41 112 76 633 RHCs 307 119 100 58 584 BHUs 2,494 781 1,135 432 4,842 Total 4,342 1,306 2,019 1,245 8,912 Source: Economic Survey of Pakistan (2012-13)

2.6 Public Health Services

The BHUs provide medicinal and preventive services for between 5,000 to

10,000 people. Rural health centres (RHCs) provide extensive outpatient services but inpatient services are usually limited to short-term observation and treatment for people who do not require transfer to a higher-level facility. RHCs serve about 25,000 to

50,000 people and are staffed by about 30 individuals including doctors and paramedics. Each RHC has approximately 10-20 beds and X-ray, laboratory, and minor surgery facilities (34).

Tehsil (sub-district) headquarter hospitals provide limited inpatient and outpatient services. They serve between 100,000 and 300,000 people with 40 to 60 beds and other services such as X-ray, laboratory and surgery facilities. The district

48 headquarter hospitals serve between half to one million people and provide a range of specialist care in addition to basic hospital and outpatient services. They typically have about 125-250 beds.

All health services in the district are controlled by the District Health Officer

(DHO). All Tehsil Headquarter hospitals managers and first-level care facilities give their reports to their DHO. Consultant surgeons head the district headquarter hospitals who along with DHOs report to the Director General of Health at the provincial level.

The Provincial Secretary of Health is directly responsible for the tertiary care hospitals.

2.7 Private Health Services

There are more than 20,000 small office-based clinics for general practitioners in the private health sector in Pakistan. These include more than 300 maternity homes called Mother and Child Centres (MCH), about 340 dispensaries for outpatient primary healthcare facilities, and more than 450 diagnostic laboratories. With regards to small to medium sized hospitals, there are also more than 500 units with 30 beds per hospital.

These small to medium-sized hospital generally cater to low risk patients and are equipped with basic obstetric, surgical and diagnostic procedures. Most of these private health services are concentrated in urban areas. Large rural hospitals tend to be run by nongovernmental organisations (NGOs) [Table 9].

Table 9: A Comparison of the Public and Private Sector Health Care System in Pakistan (39) PUBLIC SECTOR PRIVATE SECTOR

Facility Number Facility Number

Hospitals 1,270 Hospitals 106

Dispensaries 5,382 Small hospital 520

RHCs 550 General practitioners 20,000

BHUs 5,404 Maternity homes 300

MCH Centres 696 Dispensaries 340

TB Clinics 285 Diagnostic Laboratories 450

Total hospital beds 101,173 (1,786 NGOs 254 persons / bed)

Table 9 shows the summary of health services and human resource in health sector of Pakistan. 2.8 The Expanded Programme on Immunisation (EPI) in Pakistan

Pakistan became part of the accelerated health programme in 1983 after 5 years of EPI launch by the WHO in 1978. The programme was evaluated and commended by an international commission in 1984 after it successfully achieved an immunisation coverage of about 95 percent of children between the ages 2 to 5 years old (40-41).

Initially, only smallpox, BCG, diphtheria, tetanus and pertussis were included in the vaccination programme. The programme gradually evolved and six diseases were chosen on the basis of high burden of disease and the availability of well-tried vaccines at an affordable price. These currently include tuberculosis, diphtheria, neonatal tetanus, whooping cough (pertussis), poliomyelitis and measles However, following these initial successes, the coverage for vaccination has decreased dramatically. Current estimates suggest that immunisation has dropped to 54 percent, with the coverage for vaccination of pregnant women with tetanus toxoid even lower. This indicates weaknesses in the 50

EPI programme. There is evidence that the decrease in vaccination coverage may result in poor BCG vaccination in KPK (41).

In response to the situation, the government took a number of steps to further improve EPI services by accessing funds from the Global Alliance for Vaccines and

Immunisation (GAVI). Additional resources were targeted at training additional staff to facilitate vaccination, improved surveillance, introduction of new vaccines (e.g. hepatitis B vaccine) and provision of cold chain equipment (42). The Pakistan EPI has also taken up the challenge of eradicating poliomyelitis by immunising every child in the country; however, it is still unsuccessful in achieving polio eradication. According to the WHO, 81 cases of polio were recorded in Pakistan during 2013. The WHO imposed travel restrictions necessitating polio vaccination for all Pakistani residents intending to travel abroad (43).

Table 10: Pakistan EPI Childhood Immunisation Plan (42) Age Antigen At Birth 6 weeks 10 weeks 14 weeks 9 months TB BCG Polio OPV 1 OPV 2 OPV 3 Diphtheria ,tetanus, DTP 1 DTP 2 DTP 3 pertussis Hepatitis B HB 1 HB 2 HB 3 Measles, Mumps, MMR Rubella

2.9 National Tuberculosis Control Programme

In 2003, the World Health Organisation (WHO) re-emphasised the importance of TB as a global emergency declaring that tuberculosis (TB) was a threatening disease for mankind (40). Pakistan was identified as one of the countries with significant problems in relation to TB. The government of Pakistan declared TB as a national emergency after recognising it as a major public health problem. Pakistan recently 51 passed a resolution through its national assembly reiterating its commitment to control and subsequently eliminating the disease as one of its Millennium Development Goals

(44). Thereafter, Pakistan has implemented a programme aimed at facilitating the prevention of TB. The National TB Control Programme Pakistan (NTCP) was created and is now running as the largest health programme in the public sector across Pakistan.

In this regard, the Ministry of Health initiated the National TB Programme to curb the spread of TB through enhancing public and private sector health facilities (1).

The overall objective of NTCP is to reduce mortality, morbidity and disease transmission so that TB is no longer a public health problem. NTCP and the Ministry of

Health developed a comprehensive strategy for control of tuberculosis in the country with the focus on expansion and consolidation of Directly Observed Treatment Short

Course (DOTS) services in the country. The national programme was successful in establishing a country-wide network of laboratory services, providing free diagnostic and treatment services at 5,000 public health facilities and engaging the private sector to further enhance the quality of TB services. The programme was able to achieve the

Millennium Development Goals (MDG) output indicators by detecting 70% of TB cases in the country and treating 85% of them, suggesting that Pakistan was now well on track to reach MDG targets by 2015 (45).

The public education campaign associated with this programme of TB control was extensive. Posters identifying "Free services for diagnosis and treatment of TB available here" were affixed on all TB centres across the country in order to empower

TB patients to demand their rights to free treatment and diagnosis. And to ensure proper monitoring, a well-coordinated surveillance system was put in place for prompt response to drug or chemical shortages in these centres. It is because of interventions such as these that the case detection rate (CDR) and treatment success rate (TSR) for

TB have increased from 70% to 74% and from 88.% to 91.%, respectively. CDR and

TSR are the two key indicators that have been used to measure the burden of TB (46).

2.10 Summary

Chapter 2 has explored the geography, health services structure and the role of national TB control programme in the control of TB in Pakistan. Chapter 3 presents a detailed literature review exploring risk factors of TB with particular relevance to

Khyber Pakhtunkhwa.

CHAPTER 3: LITERATURE REVIEW

3.1 Introduction

Mycobacterium Tuberculosis (TB) is a serious public health threat globally and particularly in developing countries (47). Among the 22 high burden countries, Pakistan ranks 5th with a diagnosis of about 420,000 new cases every year (48). Globally, case notification is higher in males as compared to females but in Khyber Pakhtunkhwa

(KPK) the rate of notification is much higher in females (6). There is similar data reported from Afghanistan, where 68.50% new sputum positive patients were female

(49).

To explore the causes of gender differences in notification of TB, it is essential to identify potential risk factors, evaluate patients and physician's knowledge of, diagnosis of and treatment of TB. Existing literature can help us to understand and interpret the results of different studies and utilise these findings to inform the theoretical basis of my proposed research.

3.2 Aim of the Literature Review

A literature review in the context of this project is aimed at assessing the factors associated with the increased notification of TB in women and exploring the range of methodologies available to investigate and explore the risk factors.

The review also contributes to developing a theoretical base for my research identifying key areas for data collection (Social and Epidemiological Risk factors,

Patients and Physicians Knowledge).

3.3 Methodology of Literature Review

3.3.1 Systematic Review

Comprehensive analysis of the existing literature is required for identification of possible risk factors and evaluation of knowledge, perception, diagnosis and treatment of TB. To examine the issue of higher notification of TB in females, I looked at literature from a variety of sources (Global, Regional, National and Local). This helped me to identify global and regional differences - and allowed me to compare national and local studies - so that I could identify the different variables that would be used for the data collection.

3.3.2 Search Strategy

Ultimately there are three types of time dimension in research: cross-sectional

(collecting data at a single point in time), longitudinal (more than one point in time) or retrospective (asking about previous data and comparing it to now). For the literature review I chose the retrospective method for looking back at what was previously known so that I could then go on to compare it to what I found in the interview stage. I was later able to bring them together in my discussion of results. Relevant literature was identified through covering key theoretical papers and review articles on risk factors for

Tuberculosis including gender issues. Databases accessed were MEDLINE, EMBASE and Global Health, through the OvidSp interface focusing on literature published between January 2000 and February 2015. These dates were in a convenience range, as they marked the National TB Control Programme range up to the end date of my research project. The websites of specific journals, including the Journal of Infectious

Diseases and American Journal of Public Health (AJPH) were used to locate references cited in articles (mostly review articles). Additionally, the websites of the World Health

Organisation (WHO), the National Tuberculosis Control Programme (NTCP) and the 55

Department of Health, Pakistan, were used to find related reports. Finally, theses and documents held by universities and important unpublished review documents were also searched for grey literature. Data were extracted from the selected 85 research articles for tabulation, comparison and synthesis.

The search strategy sought material evaluating epidemiology and gender differences, risk factors, patient knowledge and general practitioners’ knowledge of TB in Pakistan and Khyber Pakhtunkhwa. Key words “Tuberculosis or TB”, “risk factors”,

“adult female or females”, “Pakistan” and “Khyber Pakhtunkhwa or NWFP” were used to search at the national level. I also used key words “China”, “India”, “Iran”, and

“Afghanistan” for available studies to compare the situation of the disease at the regional level [Table 11].

3.3.2.1 Search Strategy for Objective 1

The first search aimed to identify papers and reports describing the epidemiology of Tuberculosis in Pakistan and Khyber Pakhtunkhwa previously known as NWFP. Key words “Tuberculosis or TB”, “Epidemiology”, “Pakistan” and “Khyber

Pakhtunkhwa or NWFP” were used for the specific search at national level. I also used key words “China”, “India”, “Iran”, and “Afghanistan” for available studies to compare the situation of the disease at the regional level. From the above search strategy, 34 potentially relevant published articles, globally, regionally, nationally and locally were identified [Table 11-Table 12].

3.3.2.2 Search Strategy for Objective 2

The second search sought material evaluating risk factors for TB in adult females in Pakistan and Khyber Pakhtunkhwa. Key words “Tuberculosis or TB”, “risk factors”, “adult female or females” “Pakistan” and “Khyber Pakhtunkhwa or NWFP” were used to search at the national level. I also used key words “China”, “India”, “Iran”, 56 and “Afghanistan” for available studies to compare the situation of the disease at the regional level. From the above search strategy, 24 potentially relevant published articles, globally, regionally, nationally and locally were identified [Table 11-Table 12]

3.3.2.3 Search Strategy for Objective 3

The third search sought material evaluating the knowledge, attitude, awareness and treatment seeking behaviour in TB patients. Using the terms “tuberculosis”, “TB”,

“Knowledge”, “Attitude” and “Patient” as key words yielded 11 potentially relevant published articles globally, regionally, nationally and locally [Table 12].

3.3.2.4 Search Strategy for Objective 4

The fourth search was to address the awareness, opinion and management of

TB amongst family physicians. Using the terms “Tuberculosis”, “TB”, “Knowledge”,

“Attitude,” “doctors”, “General Practitioner”, “GPs” and “Family Physician” as key words yielded 16 potentially relevant published articles globally, regionally, nationally and locally [Table 12].

Table 11: Search Strategy and Terms Justification - To Search Terms Search Group limit results to papers Rationale

relating

Tuberculosis/TB China, India, Iran, Afghanistan Regionally AND “Epidemiology of TB” OR Epidemiology Pakistan OR Nationally Tuberculosis/TB

AND NWFP OR Khyber “Risk Factors for Adult TB Locally Risk Factors Pakhtunkhwa Patient” Adult Female/s In choosing these terms I China, India, Iran, Afghanistan Regionally endeavoured to Tuberculosis/TB OR “Knowledge, attitude, awareness AND fully define the terminology for and treatment seeking behaviour Knowledge OR Attitude OR Pakistan OR Nationally among the TB patients” Awareness OR Patients the review and NWFP OR Khyber maintain Locally Pakhtunkhwa consistency in Combined with (AND) with Combined interpretation China, India, Iran, Afghanistan “Tuberculosis” Regionally AND OR “Assessment of Awareness, “Knowledge” OR “Attitude” OR manner and management of TB Pakistan OR Nationally “Awareness” OR “GPs” OR amongst Family Physicians” “Doctors” OR “Family NWFP OR Khyber Locally Physicians” Pakhtunkhwa

Table 11 on page 58 shows in detail what, search terms were used for each task with a justification and rationale. Searches were limited to references about empirical research or review articles.

The results of all these searches were exported into a database using Reference

Manager. Then, the titles and abstract of each reference were read carefully to decide which papers should be reviewed critically. Therefore, the following criteria were used to identify relevant papers for review:

3.4 Inclusion Criteria

 Empirical research or a systematic review relevant to the epidemiology

and risk factors for TB.

 Articles published in English

 Knowledge, Attitude, Awareness and Treatment seeking behaviour in

TB patients-related papers.

 Assessment of learning, approach and management of TB amongst

family physicians-related papers.

3.5 Exclusion Criteria

 The research involved a case report or case series

 Involved bovine Tuberculosis (BTB) or did not differentiate TB from

BTB infection.

 Research involved individuals at higher risk of TB because of

underlying disease (for example, research on subjects with lung cancer

or HIV/AIDS).

3.6 Summary of Literature Selection

The systematic review identified 4,630 screen citations. Of the total, 2530 studies were yielded by MEDLINE search, EMBASE search yielded 1155 and

GOLBAL HEALTH 730 articles. Additionally, from the websites of WHO, NTCP,

MOH and online journals (JID, AJPH) a further 250 articles were identified.

3067 articles were excluded as not relevant to the research area because they did not meet inclusion criteria; for example several focused on children and young people. Further review of the articles by reading the titles and abstracts, resulted in removal of a further 749 articles as not relevant. After reading full text articles, 154 studies were considered eligible on the basis of our inclusion criteria. Finally, after removing the duplicate articles 85 studies were identified for the literature review

[Figure 7] (PRISMA Check list available as appendix 14.8 on page 235)

Figure 7: PRISMA Flow Diagram

Formulate Research Title, Aims and Objectives, Research Questions and Hypothesis

Database Searches MEDLINE, EMBASE, and Studies identified GLOBAL HEALTH. (N=4,630) WHO, NTCP, MOH, JID and AJPH (3,5)

Research Articles screened by Title and Abstract Studies excluded N=1563 (N=749)

Full Text Papers Studies excluded N=814 (N=660)

Articles considered eligible Duplicate articles for Review (N=69) N=154

Articles finally selected for systematic review Epidemiology= 37 Risk factors =25 Patient Knowledge= 9 GP Knowledge =14 N=85

Table 12: Findings from the Different Search Strategies Regional Task Database Global National Local Total China India Iran Afg*

Medline 17 6 7 5 3 9 3 50

Embase 15 7 6 3 1 4 5 41 Global Health 13 6 4 2 2 7 3 37 WHO/NTCP/Other 10 2 1 2 1 7 3 26 Total Searched 55 21 18 12 7 27 14 154

Search Strategy Strategy Search Articles Total After 27 12 10 7 5 16 8 85 Duplication 1 TB Epidemiology 11 4 3 3 3 5 5 34 Risk Factors for TB 2 6 3 3 2 2 6 2 24 Patients Articles Related to 3 Knowledge of TB 4 2 2 1 0 2 0 11 Patients Publication Related 4 to Review of GPs 6 3 2 1 0 3 1 16 Knowledge about TB Afg*; Afghanistan

The four different search strategies identified the following relevant papers: 34 on the epidemiology of TB, 24 on the risk factors for TB, 11 on patient knowledge about TB and 16 on GPs’ knowledge about TB.

Table 13: Summary Table of the Literature describing Gender Differences of TB in Pakistan Author Name Study Design or Study Subjects and Aim Key Findings Critical Observations (Year) Methodology Recruitment Methodology Khadim et al Factors affecting TB Cross Sectional 100 (M=56, F=44) Health awareness, easy accessibility to health No statistical methodology mentioned in the (2003) control: Decision making Study Newly registered TB patients in care facility qualified heath care provider study, however detail diagrammatic (50) in the household level the Rawalpindi TB Centre play important role National TB Control presentation makes the study unique between Jan to Feb 2002 included in the study. A questionnaire filled by every 5th adult participant in the serial. Highlights TB burden and No study design no Budget for research activities in Pakistan is This is an article which provides Vermund et al importance of research in very low comprehensive exploration of the diseases (2009) Pakistan and importance of research in the region (19) especially in Pakistan

Ullah et al (2008) To identify trends in the Prospective study 525 (m=176, f=349) Women are at higher risk of Extra Pulmonary Very well written paper .The results provide (51) Extra Pulmonary TB Study subject recruited as TB than male in NWFP some support for our study hypothesis incidence in NWFP. per submission of the biopsy for TB To identify age related Prospective study 67 (M=35, F=32) Dyspnoea in elderly and upper lobe Statistical software or methodology used is Rizvi et al (2003) differences in the clinical infiltration are common symptoms of TB unclear, however clear and detail statistical (52) presentation of TB analysis done.

Is the Ministry of Health Descriptive Study 12 (M=6, F=6) Information and communication system of Although, sample size of the study is very Israr (2003) fully prepared to the Health care system of Pakistan is very small it explored very important area of (53) implement an effective poor for effective control of TB research (DOTS). DOTS programme in Pakistan? Prevalence of Pulmonary Cross Sectional 386 male Results showed that prevalence of TB among Single centre study and limited to male Shah et al (2003 TB in Karachi Juvenile Study the prisoners was 3.9% patients only. Insufficient training of the Jail, Pakistan technicians with language barriers (54) Targeting the neglected group of the community where all the risk factors present in single premises.

Recent trends in the Descriptive Study 150 (M= 77, F=73) Atypical pattern of TB increasing diagnosing Study adopted strong descriptive study Rao and Sadiq, radiological presentation All newly diagnosed TB delay design. But limited to single centre (2002)(55) of Pulmonary TB in cases with x ray chest were Pakistani adults included in the study. 63

Muynck et al TB Control in Pakistan: Literature review Literature review NTP facing very serious technical and A retrospective study, limited (2001) Critical analysis of its managerial weakness, with strong political to previous research articles implementation interventions. Comprehensive critical analysis of the NTCP Research based activities should be and ministry of health in Pakistan promoted and importance should be given to those areas where data gaps are present NTCP Report, Reporting annual Annual Report National report A total of 11362 New Smear Positive, 4816 Provincial data of TB regarding the new cases 2006 (56) activities of National TB New Smear Negative, 1063 Extra Pulmonary and gender distribution were missing. Control Programme and 148 Re-treatment were notified in 2006. Pakistan achieved 100% DOTS coverage in 2005 and sustained. Treatment Success Rate was reported 85% in 2006. Additionally, 700 health professionals including doctors, medical and laboratory technicians were trained NTCP Report, Reporting annual Annual Report National report Incidence of TB in Pakistan 177/100,000. TB On the basis of case notification Pakistan 2007 (57) activities of National TB responsible for 5.1% of the total national ranks 6th amongst the high burden countries Control Programme morbidity burden in Pakistan. globally in 2007. NTP Report, Free TB Pakistan: Current Annual report National report NTP achieving its aims and objectives as per Annual report of NTCP based on population 2008 (21) status and the way planning. Incidence of TB in Pakistan of 1998 where there is huge increase in the forward increased from 177/100,000 to 181/100,000 last 10 years. in 2008. But the global ranking of Pakistan Well written report highlights the key improved from 6th to 8th, which reflects the components in TB epidemiology and control NTCP activities of fighting against TB in in Pakistan. Too many figures and discussing Pakistan the achievements of the programme but didn’t report weak areas of the programme. NTCP Report, Reporting annual Annual Report National report A total of 267,491 new cases, including A detailed report having the data of all four 2009 (44) activities of National TB 101,887 sputum positive were reported in provinces and reporting that more female Control Programme 2009. Incidence rate was 181/100,000. patients were notified in Balochistan. But the report missed the gender notification of TB in KPK which was also higher in females. NTP Report, UPSCALING Annual Report National report Incidence of TB in Pakistan 231/100,000 A detailed report of TB burden in Pakistan (2010) TUBERCULOSIS 269,290 new TB cases (all types) were but data regarding foreign nationals with TB (45) CARE AND CONTROL notified during 2010. is missing. The report showed that female Epidemiology of TB in Of these, 104,263 cases were SS+, these notification was significantly high in some KPK figure are higher as compared to last year. Balochistan but missed KPK in reporting. NTCP Report, Annual report National report Incidence rate was reported 177/100,000. This was the first comprehensive and detailed 2012 (2) Pakistan jumped to 5th in global ranking with report of NTCP. More TB patients were 55% of total burden of TB disease in EMR reported as compared to 2011. Report also 64

region. A total of 269,265 TB patients were highlighted the issue of multi drug resistant reported in 2012, of these 104,238 cases were TB in Pakistan. Health services in Pakistan sputum positive. also reported in 2012. The report also showed the summery of notified cases from 2001 to 2012. NTCP Report, Annual Report National report Pakistan jumped to A total of 298,446 new In-spite of 100 % DOTS coverage achieved 2013 (1) cases notified during 2013, compared to in 2005, still 27 deaths were reported in 2013 273,097 in 2012, with an incidence of from a disease which is curable. Pakistan 264/100,000. Pakistan ranks 5th globally in share about 65% of total TB burden in EMR term of TB burden. region in 2013, comparing to 55% in 2012. WHO Report Gender and Tuberculosis Report of research Multi Countries Study Males are at higher risk than females in India Comprehensive study exploring the burden of (2006) and training in and Bangladesh (neighbouring countries of TB in the region, but surprising to see that Tropical diseases Pakistan) two high burden countries in the region (China and Pakistan) were not reported , WHO Report, Global Tuberculosis Annual Report Worldwide population Report showed that Pakistan one of three Report provides key information about TB 2007 (18) Control countries spending less half of the available worldwide for the year 2007. SURVEILLANCE, funds on TB in 2007. PLANNING, A total of 1.6 million people died of TB, FINANCING including 195 000 patients infected MDR-TB a serious problem in HBC WHO Report, Global Tuberculosis Annual Report Worldwide population A total of 2.5 million new sputum positive The report was based on data from 2006 2008 (58) Control cases were reported in 2008. Survey of mainly, not 2007 or 2008. However, it still it SURVEILLANCE, knowledge, attitude and practice (KAP) were has a lot of new information about the global PLANNING, carried in 13 high burden countries. Pakistan TB burden. FINANCING ranks 8th globally for TB burden. WHO Report, Global Tuberculosis Annual Report Worldwide population A total 9.27 million new cases notified in A detailed analysis of the global TB burden 2009 (15) Control 2009, compared to 9.24 million in 2008. 80% and provides key information for planning Epidemiology, Strategy , of TB occurs in 22 high burden countries, and policy making. Financing including Pakistan. In 2009, total budget available in 22 HBCs was US $ 2.2 billion for TB. WHO Report, Global Tuberculosis Annual Report Worldwide population A total of 9.4 million new TB cases reported Data reported in WHO TB report for 2010 2010 (14) Control in 2010. Pakistan secured funding for survey was not as comprehensive as reported in in 2008 but due to security concerns the 2008, 2007 and 2006. survey still has not been implemented. WHO Report, Global Tuberculosis Annual Report Worldwide population A total of 8.8 million TB cases reported in The report showed the TB burden in detail 2011 (20) Control 2011, these figures were less as compared to globally. Most importantly, the number of 2010 (9.4 million). China and India both cases decreased in 2011 as compared to 2010. share 40% of the total TB burden. 65

WHO Report, Global Tuberculosis Annual Report Worldwide population In 2013, 8.6 million TB cases reported A detailed report for the global TB burden, 2013 (13) Control globally, where about 450 000 patients which highlights MDR-TB and its threat in developed MDR-TB and of those about 170 developing courtiers. 000 people died. National survey on drug resistant carried out in Pakistan. In 2013 WHO reported 1:1 Male to female ratio for TB in Pakistan. WHO Report, Global Tuberculosis Annual Report Worldwide population In 2014, the numbers of notified TB patients The report described reliable data on the 2014 (12) Control were increased to 9 million, compared to 8.6 global TB burden. million in 2013. TB treatment success rate was reported 86% for new TB patients in 2014.But still 1.1 million people died from TB alone. WHO Report, Global Tuberculosis Annual Report Worldwide population A total of 9.4 million TB cases were reported The 2015 global report for TB is the first 2015 (59) Control 20th Edition in 2015, of those 5.4 million men, 3.2 million report that has a separate section (Section 4) women and 1.0 million were children. WHO on gender differences in TB, which provides recommends that detection and treatment reliable data for policy making. gaps must be addressed to reduce the TB burden. About 190,000 people died from MDR-TB in 2014. The report highlights that national scurvies identified gender differences in TB and should be investigated Nassaji et al 2014 To evaluate Cross Sectional A total of 226 patients, of More female patients were reported as A well written Study with abstract, Afghanistan (60) demonstration of acid fast Study those 46.5% (105) were compared to female. background, method, results and discussions. bacilli in tissues section male and 53.5% (121) Limitations of the study were missing. showing typical female were diagnosed with histopathological features TB. of tuberculosis in patients with EPTB using ZN staining method. Fader et al, 2010 To describe the Retrospective 118 (male to female ratio = The average age of males at the time of A study which showed that female patients in Afghanistan (61) epidemiology of EPTB in Study 2:1) diagnosis was 32.2 years and of females 31.1 Afghanistan were at a higher risk for TB than Afghanistan. years. males.

Khan & Laaser, Burden of Tuberculosis in Tuberculosis Health services review Incidence of TB in Afghanistan 278/100,000. The paper based on the data collected in 2002 Afghanistan: Update on a control Number of death reported in 2001 15,000. Pakistan not in Afghanistan by NGO working Afghanistan (62) War-stricken Country programme Male to female distribution 30/70. DOTS for the control of TB in Afghan refugees Ibrahim coverage 10%. camps.

Roya Alavi- To reveal the gender Cross Sectional A total of 2,284 of TB Of the total (2,284), majority 1205 (f= 726, Figures in the study showed that all form of Naini, 2007 differences in treatment of Study patients were registered, of m= 479) were sputum positive, 552 (f= 331, TB were more common in females as Iran (63) TB those 1375 (60.2%) females m= 221) sputum negative and 527 (f= 318, compared to males. It is a large sample size and 909 (39.8%) were m= 209) were reported extra pulmonary TB. study but the study design and recruiting males. methodology of patients was not clear. Chaha , 2005 A review of Tuberculosis Review Article Medline, Medscape and TB situation was reported worse in rural Gender distribution of TB was not reported. India (64) epidemiology in India. AIDSLINE databases were areas. India accounts 20% of the global TB searched. Reports also burden with estimated 1.8 million TB cases collected from NTP and every year. other websites Viney et al, 2011 The Epidemiology of Retrospective A total of 1544 TB cases Of the total, 73% (1105) were pulmonary and Multi country study with reliable data, which Global (65) Tuberculosis in Pacific study were notified (male/female 27% (409) were extra-pulmonary. Of all provides key information about TB globally. Island Countries and = 1.0 /0.8 pulmonary TB cases, 52% (575) were sputum Territories positive and 48% (530) were negative. Kakchapati et al, Model incidence of TB Retrospective A total of 198,719 cases Epidemiological findings demonstrate that Study showed the gender difference in the 2010 between 2003 and 2008 in study reported between 2003- tuberculosis incidence rates are higher for incidence of TB in Nepal. Nepal (66) Nepal. 2008, of those 127,979 males compared to females. Rural areas were (64.4%) were male and also reported with higher incidence compared 70,740 (35.6%) were to urban. female. Kadivar et al The study aim to evaluate Descriptive A total of 1026 of new Results showed that 66.8% of cases were The occurrence of TB in Afghanistan is 2001 the epidemiology of TB in Epidemiological sputum positive (NSS +ve) aged 15-44 years. Pulmonary TB (74.9%) higher in females as compared to male, but Iran (67) Afghans in southern Study patients were reported, of was more common than Extra Pulmonary the results of the research reported more Islamic Republic of Iran. those 371 where were (24.1%). The incidence of TB was 92.2 ( males than females, which is not similar to Afghans. (male= 56%, 2000), 88.2 (2001) and 63.5 (2002) per the results in Afghanistan. female= 44%) 100,000 population Dogar et al, 2012 The aim of the study was Cross Sectional Study carried out on NTCP A total of 571,958 new smear sputum (NSS) This study reported a reliable data and Pakistan (6) to see whether gender Study notified cases between 2001 positive cases were reported between 2001 initiated on the basis of our preliminary study differences are evenly to 2010 and 2010. Of those 51% (292,551) were carried out on the 10 years NTCP data in distributed across the males and 49% (279,407) were females. KPK in 2011. Results of this study confirmed country or vary by However, the notification of TB cases in the findings of our study for objective one geographies, to enable females was higher in KPK (MFR= that female patients are at higher risk for TB effective targeting of TB 0.70:1.00) and Balochistan (MFR= in the province of KPK, Pakistan. control strategies. 0.68:0.78), compared to Punjab (MFR= 1.12:1.00) and Sindh (1.27:1.00).

Table 14: Summary Table of the Literature of Risk Factors for Tuberculosis

Author Study Design or Study Subjects and Aim Key Findings Critical Observations Name (Year) Methodology Recruitment Methodology Akhtar and To identify risk Cross Sectional Study 385 (m=205, f=180) Household contact of TB exposure is Study is limited to single centre and methodology Rathi, 2009 factors for TB Cases identified from TB clinic more significant and need attention for recruitment is also unclear. However research (68)) register in Umerkote highlighted an important issue of house contact of TB patients in Pakistan. Soomro and To Compare Factors Cross Sectional Study 100 (m=62, f=38) Significant variations observed between It is a study with small sample size and limited to Qazi, 2009 Associated with All patients with relapse were male and single centre. However, the research results are (69) Relapsed TB in included in the presented to out females valuable focusing on social and potential risk Male and Females patients department of Civil factors, such as overcrowding and HB level Hospital in Hyderabad. Aziz et al, To assess the risk Cross Sectional Study 200 participants recruited as Results showed that household contacts Methodology for selection of study subject is 2008 (70) for family member household’s member having of TB patients are at higher risk than missing. Similarly gender of participants was not of TB patients active Pul. TB. non contacts described in the study. Furthermore, Lahore is a contacts as large city so selection of house hold is unclear. compared to non Method for statistical analysis described in detail. contacts. Javaid et al, To identify drug Cross Sectional Study 199 (m=53, f=66) Women are at higher risk than male in The study does not explain the ratio of drug 2008 (71) resistance to anti TB All newly diagnosed TB NWFP for MDR resistance by gender. drugs in NWFP, patients in 7 centres of NWFP Multi centres study with significant results Pakistan. Hussain et al, To identify risk Cross Sectional Study 425 High level of latent TB infection need Gender aspects of the participants ignored in the 2003 (72) factors for latent TB 164- Peshawar urgent attention from NTP study. While some of the participants from risk infection in 83- Haripur group were found missing from screening. prisoners of NWFP. 90-D.I. Khan It is a multi-centres study covering the entire 54-Kohat province. The study also recommends further research for the diagnosis of TB in high risk groups. Rizvi et al, To differentiate the Prospective Study carried 67 Young COPD is more common risk factor in It is unclear that which statistical software or 2003 (52) pattern of out from Dec 1999 to 36 Elderly elder patients methodology was used in the study. However, the pulmonary TB in May 2000 The two groups divided below results were presented clearly younger and older 60 for younger and above 60 for patients older and recruited all the study subjects in this age groups between the study duration

Liefooghe To identify different Semi structured 563 (44% = male, 56%=female) Majority of the patients were in the It is a study of small sample size, limited to single and Muynck factors which questionnaire based study Patient recruited on the basis of most productive age (15-45), Illiterate district. , 2001 (73) Tuberculosis having both qualitative sputum positive test and above patients have high defaulter rate. Study design not clear, abstract have no results management and quantitative approach 15 years of age Irregular use of anti TB drugs is a and organised as like a conclusion significant risk factor Sharma et al To examine the Survey A total of 4,91,000 of covered Of the total 74% were belong to rural Clear study design was not mentioned and 2010, trend in gender gap population nearly half were and 27% were from urban area. recruiting of ample size was unclear as well. India (74) in the prevalence of female. Significant gender difference was noted However, the study showed gander differences in TB over the year in rural areas 98/ 100,000 v rural and urban areas and relevant to our study as 176/100,000 (P<0.05). 75% of the KPK population living in rural areas. Davies, Risk Factors for Review article Literature review Malnutrition, young age, Well search literature review, identified main risk 20005 Tuberculosis socioeconomic, indoor air pollution, factors responsible for Tb disease. In our study (9) alcohol and health system issues were we also assume some of these risk factors for identified as main risk factors for TB. higher notification of TB in KPK. Sabawoon To explore Sex Retrospective Study of A total of 9,949 NSS +ve were Female patients in Afghanistan who Retrospective study based on the data of NGOs and Sato, Difference in TB records reported, of those 31.5% were aged 15 through 44 years, were and TB control programme The geographical and 2012 Tuberculosis in (3,131) were males and 68.5% twice more likely have TB than males. cultural situation of Afghanistan is similar to the Afghanistan Afghanistan (6,818) were females. Similarly, females in Afghanistan stay province of KPK, where we conducted our (49) for long time inside in homes with poor research. Therefore findings of this study were ventilation and lighting (i.e., away from very relevant and important to our results sunlight that quickly kills tubercle bacilli)

Table 15: Summary Table of the Literature on Patient’s Knowledge about Tuberculosis Author Name Study Design or Study Subjects and Recruitment Aim Key Findings Critical Observations (Year) Methodology Methodology Mushtaq et al, To Explore inequities in Cross sectional 1080 (Urban=432, Rural=648) In the rural population Unavailability of female staff for interviewing females. 2011 (75) knowledge, attitude and Study Male=76.2% knowledge regarding TB Repetition of study previously published with same results practices regarding TB among Female=23.3% was deficient. in 2010. the urban and rural Two out 35 distracts in Punjab However, it is important study, comparing TB in both populations. were randomly selected. Five urban and rural communities. Union council two from urban and three from rural were further selected randomly. From each Tehsil two villages selected for interviewing 18 indivisibles Samera et al, To assess the level knowledge Hospital based 301 (m=160, f=141) Awareness about TB It is study limited to only two centres 2008 (76) of TB patients regarding TB cross sectional Hospital based study, all TB was very deficit in TB study patients attending the setting from patients contributing to May to July 2006 recruited for the delay in diagnosis of TB study. Khan et al, To assess the knowledge of Cross Sectional 170 (m=92, f=78) Misconceptions about Limited to a single city only, but covered both the 2006 (77) patients with TB about their Study Patients above 16 years having TB TB were widespread in Government and Private health facilities. disease and misconceptions now or in past were recruited for Pakistani Patients. regarding TB the study, seeking treatment in two hospitals. Ali et al, 2003 To explore the level of Descriptive cross 203 (m=131, f=72) Knowledge about TB in Findings limited to only one centre where the awareness (78) knowledge regarding TB sectional survey both patients and their level is high as compared to other areas amongst patients and their based study family was deficit and Health education should be a part of such studies in future families larger population based TB patients and their families members were the most studies should be affecting people, so important area for research conducted Agboatwalla To assess the knowledge and Cross Sectional 754 (m=385, f=369) Overall Knowledge of Selection of urban and rural sites in methodology was not et al, 2003 attitude about TB in rural and descriptive study One urban and rural site in Karachi TB was deficient in rural mentioned. However, the approach of interviewing the (79) urban community of Sindh were selected randomly for females. study subjects, male by male and female by female was Provence of Pakistan interviewing 100 household each very appropriate. Similarly, is very important research focusing on gender perspective of knowledge about TB

Table 16: Summary Table of the Literature on Physician’s Knowledge about TB Study Design Author Study Subjects and Recruitment Aim or Key Findings Critical Observations Name (Year) Methodology Methodology Ahmed et al, To assess the knowledge, Questionnaire 22 PPs High gaps in the diagnosis It is single distract study limited to male private General 2009 (80) attitude and practices of based survey Participant recruited from a TB and treatment of TB practitioners only. Similarly, clear study design is not private GPs regarding TB DOTS workshop having medical through DOTS observed mentioned in methodology.. However, this is the first survey DOTS graduation in the rural area of Sindh, Pakistan Shehzadi et To assess the knowledge of Cross 88 GPs were recruited from Gilgit, Sever deficiencies were The methodology for sampling and recruiting of GPs is not al, 2005 (81) GPs in NWFP and Northern Sectional Hunza, Skardu, Abbottabad and seen in the treatment of clear in the study. Similarly, the study limited to only male Areas of Pakistan regarding Survey Peshawar TB by GPs in NWFP. GPs. However, the study highlights the important area of the management of TB research (GP’s Knowledge) in six distracts of NWFP. Hussain et al To evaluate the knowledge Cross 53 GPs recruited who fulfilled the Research observed A study of small sample size, limited to male PPs, however (2005) (82) of PPs regarding TB Sectional inclusive criteria for the study significant deficit of the statistical analysis of data and the tabulated presentation management Study knowledge regarding TB is very good. Most importantly, it was an unique research management amongst PPs strategy for assessing the knowledge of GPs by experimental approach by acting as TB patient Shah et al, To determine the knowledge Descriptive 245(m=197, f=48) The study showed that the Study limited to only 2 cities of Pakistan. 2003 (83) of private GPs towards Cross FPs were selected from Rawalpindi knowledge and practices Majority of Pakistani population were receiving treatment in diagnosis, treatment and sectional and Lahore cities of Punjab, Pakistan, of GPs were not in line Government Hospital, so it is the strength of the study that follow up of pulmonary TB Survey. working in Govt health care facilities with National TB Control research performed in public health care. Similarly, study in 2 large cities of Pakistan. with basic degree in medicine. Programme guidelines. covered all the basic components of TB management i.e. diagnosis, treatment and follow up. Khan and To assess the practice of Cross 120 GPs out of 141GPs participated Some of the Doctors in the This is also a single centre study limited to only male Hassain 2003 prescription of doctors sectional in the study private teaching hospital doctors. (84) working in the private study prescribing TB drugs teaching hospital of Pakistan below recommended dose. Khan, 2003 To evaluate the knowledge Questionnaire 120 GPs It has been found that the Study covered only a single city of the country with no clear (85) and practices of private GPs Survey All the GPs working in Karachi were knowledge and practices study design. regarding the diagnosis and invited to join a workshop on health among Private GPs were Questionnaire of the study described in detail with important treatment of Pulmonary TB. practices and during that workshop unsatisfactory variables and meet the gender aspect of research. However, it questionnaire distributed and is not mentioned how many male or female were included in collected for the study the study. Rizvi and To investigate the A 150 Family Physicians There are many The writing standard of the study is very bad repeating study Hussain, knowledge, attitude and questionnaire 30 FPs from each five districts of misunderstandings about questions. Similarly, reference section is not according to the 2001 (86) practices pertaining to TB based study. Karachi were selected randomly the transmission of TB standard guidelines. among family physicians. and BCG vaccinations. Focus on GPs without Post Graduate qualification

CHAPTER 4: FINDINGS OF THE SYSTEMATIC REVIEW “TUBERCULOSIS IN PAKISTAN”

Tuberculosis (TB) is one of the most rapidly re-emerging infectious diseases globally and is having a significant impact on the health burden in Pakistan. The national incidence of TB is 177/100,000 which means that Pakistan is ranked 5th in the

High Burden Countries (HBC) worldwide. Pakistan ranks first with 43.0% share of the total TB burden in the Eastern Mediterranean Region (40). National TB Control

Programme (NTCP) reported that TB contributes a 5.1% share to the overall morbidity burden in Pakistan (1)(4)(19).

The population of Pakistan is 182.2 million the sixth largest in the world - the annual growth rate is 2.7%. About 70.0% of the total population lives rurally. The province of Khyber Pakhtunkhwa is identified as an area where the notification rate of

Tuberculosis is reportedly higher in female than male - a situation which is contrary to the gender distribution of TB at the national level (6)(21). Similar data are also reported from Afghanistan, where female notification of TB was also higher in female patients

(49). Sharma et al in a survey of 90,000 households in India found that prevalence of

TB in males is higher than females (74). Similarly, data reported from the TB Control

Programme Shanghai China by Shen et al also described that the notification of male

TB patients was significantly higher than female (87).

4.1 Epidemiology of TB in Pakistan

4.1.1 Introduction

The main source of information about TB in Pakistan is the National

Tuberculosis Control Programme (NTCP) which started in 2001, with the annual reports for 2001 to 2013 describing the disease epidemiology in detail (1)(4)(21). NTCP 72 is a national organisation working with aid agencies and partner organisations for the eradication of TB in Pakistan: it also has branches responsible for TB control at the provincial level.

The control strategy involves co-ordinating the Primary Health Care (PHC) system, District Headquarter Hospitals (DHQs), Rural Health Centres (RHCs) and

Basic Health Units (BHUs); in addition there are currently 23 tertiary heath care hospitals in Pakistan working within the NTCP for the diagnosis and treatment of TB patients (36). At collecting district level, the TB Control Officer (TBCO) is responsible for reporting all newly diagnosed TB patients' data to NTCP provincial office on a monthly basis. All the TBCO have extensive training on data reporting and therefore the data is highly reliable.

According to the 2014 WHO report annual disease incidence in Pakistan is

275/100,000, which is higher than Afghanistan (189/100,000), China (70/100,000) and

India (171/100,000) (12). The notification of new TB cases of all forms was 298,446 in

2013, compared to 273,097 in 2012. In 2013 newly reported sputum positives and sputum negative cases were 111,682 and, 118279 respectively. Additionally 52,646 new extra pulmonary cases (EPTB) were registered in 2013. Multi Drug Resistant TB

(MDR-TB) is a serious public health issue with 1,570 new cases reported in 2013, compared 210 in 2010. The total burden of MDR-TB in Pakistan is about 9,900 (1).

According to WHO 0.9% of TB patients in Pakistan have co-infection with HIV (88).

Directly observed treatment short course (DOTS) coverage increased from

50.0% in 2002 to 100% national coverage in 2005. Treatment success rate has increased from 67.0 % to 91.0% in the last 10 years, which is higher than Afghanistan

(88%) and India (89%) in the region (89). Nevertheless, the mortality rate remains unacceptably high at 37/100,000. The NTP spends 19.0% of its budget on first line

73 drugs, 20.0 % for laboratory and equipment costs, and 6.6% for NTP staffing costs out of a US$65 million total budget in 2007 (21).

Vermund et al reported that Pakistan spends 0.8% of its GDP on health care, which is about 50.0% less than Bangladesh and Sri Lanka. Health care research expenditure is also very limited in Pakistan because of high state expenditure on defence (22.0%). The report identifies the need for further research to eradicate TB from Pakistan (19).

4.1.2 Age and Gender

Age and gender are the two important variables in TB research, because the age of patients plays an important role in the state economy. Globally, it is recognised that notification of TB is higher in males as compared to females (90-92). However, figures show that younger female patients in developing countries are at a higher risk of

TB than males of the same age (93). According to National TB Control Programme, TB case notification was higher in young female patients of Khyber Pakhtunkhwa and

Balochistan as compared to Punjab and Sindh, where notification of male patients was higher when compared to females (6)(94). Seventy five percent of identified TB cases occurred amongst the economically active population (15-45 years), which is a substantial economic burden at both the individual and state level (19). In a study of 386 male inmates of a Juvenile prison in Karachi, Shah et al reported that nearly two thirds of TB patients (n=239, 61.9%) were aged 17 to 18 years. Overall disease prevalence was 3.9% which was significantly higher than the estimated 1.1% in the general

Pakistani population (54). A similar study of male inmates in five prisons in the North

West Frontier Province found that of 425 prisoners, 204 (48.0%) had latent TB infection, as determined by Mantoux testing and of these just over a quarter (n=111,

26.0%) were aged 18 to 26 years (72).

Rizvi et al in a prospective study carried out in two teaching hospitals in

Karachi, reported that among 67 TB patients with a mean age of 30.6 years, just over half (n=35, 54.0%) were male: among older patients whose mean age was 65.9 year, a similar proportion were male (86). Khadim et al reported similar findings from a small cross sectional study carried out in Rawalpindi , where among one hundred TB patients, just over half (n=56, 56.0%) were male . The authors identified the need for larger population-based studies to explore risk factors for TB in Pakistan (50). Khan et al studied the awareness of 170 TB hospitalised patients about their illness – more than half (54.2%) were male (77). In a two-hospital chest clinic out-patient based cross sectional study, Samera et al recruited 301 participants of whom 160 (53.0%) were males (76). In a descriptive cross sectional study, Akhtar and Rathi reported that out of

385 study subjects, 53.2% were males and 46.8% were females (68). Somooro and Qazi using a cross sectional study design looked at 100 patients with reactivated/ re-infection

TB in the Hyderabad DHQ Hospital between August and November 2008 and found that 62.0% were males (95). In a very recent study, Mushtaq et al compared knowledge and behaviour about TB infection in urban and rural communities in Punjab province: among the 1080 participants, just over three quarters were male (75). Therefore, the evidence for the predominance of TB in men is substantial.

In contrast to these findings, Dogar et al in a recent study about gender differences in TB in Pakistan reported higher case notification for females in the province of KPK as compared to males (6). These results are reliable, as it was conducted by NTCP in four provinces of Pakistan, similar to our descriptive study in

KPK between 2002 and 2010. Ullah. et al in a recent prospective study carried out on

TB patients at the Lady Reading Hospital Peshawar, KPK also found that of 525 patients diagnosed with extra-pulmonary TB (EPTB), nearly two thirds (n=349, 66.5%) were female (51). Research data from this small study suggested that risk factors for this

75 higher female incidence was related to the male dominant culture, multi parity, poor nutritional status, lack of awareness about TB, overcrowding and the low female literacy rate in the area. Most patients (70.1 %, n=369) were diagnosed with TB in their most productive age (15 to 45 years). The authors concluded that women were at higher risk of TB than males particularly during the reproductive age. Moreover, a recent study by Khattak et al, aimed to identify the number of sputum positive cases among PTB patients, women were found to be at higher risk of TB than males in KPK. Out 104 patients, 57.40% (n=60) were female and 61.52% were between the age 20 and 50 (96).

Bassali et al reported that males were more commonly infected with TB in Egypt, Iraq and Yemen, while in Pakistan and Iran females were diagnosed in larger number as compared to males (97). These results provide evidence for the study hypothesis that females in the province of KPK are at higher risk for TB than males. Therefore, it is very important to investigate the causes of this higher notification in females, because women are mainly responsible for looking after children and families in KPK.

4.1.3 Educational Status

Educational status is the level of literacy attained by an individual by attending school, college or university. In Pakistan, there are religious institutions (Madrassa), which also provide education to the community. Literacy rates in general in Pakistan are higher in males than females (98). Several studies have reported that literacy levels are directly related to the incidence of Tuberculosis. Shah et al reported that more than half (n=239, 59.1%) of patients with TB in a juvenile prison were uneducated (54).

Similar findings were reported by Hussain et al in a cross sectional study carried out to assess the risk factors for TB in KPK, where just over half (57.0%) of their study subjects were uneducated (illiterate) with a further third (34.0%) only receiving primary education (72). Similar results were reported by Akhter et al who found that the

76 majority (n=277, 71.9%) of participants who were randomly selected from the TB patient register maintained by the Umerkot AntiTuberculosis Association clinic were categorised as uneducated (68). Similar results were reported by Somooro and Oazi who found that 58.1% males and 68.4% females were uneducated (95). Samera et al explored the relationship between knowledge and the delay in the diagnosis and treatment of Tuberculosis. Of the 301 recruited patients, just under three quarters

(71.0%, n=213) were uneducated (76). Agboatwalla et al compared literacy levels connected to TB in urban and rural settings – people were eligible to enter the study provided they had lived in the area for ≥ 6 months. They interviewed 754 participants and found that more than three quarters (78%) of respondents from rural areas were uneducated compared with only 51.0% in urban areas (79). Due to poverty, problems in transportation and cultural barriers, we would expect these results from rural areas.

Somewhat contradictory results were reported by Khan et al who found that just under a quarter (24.1%) were classified as illiterate with 10.0% having either a degree or postgraduate qualification – this is almost certainly explained by the study design which recruited nearly two-thirds (65.9%) of patients from a private hospital

(77). Similar results were found in a study by Mushtaq et al of villages in the Punjab – of the 1080 participants, the majority (n=648, 60%) were from rural areas. Their research compared knowledge and behaviour about TB infection in urban and rural communities in Punjab province, and reported that only 25.5% from rural and 16.3% from urban areas were classified as uneducated (75).

4.1.4 Socio-Economic Status

Socio-economic status (SES) is characterised by the economic and social position of an individual based on income, education and employment. Soomro and

Qazi reported that nearly three quarters (71.2%) of the patients with TB in their study

77 were classified as low income (i.e. having a monthly income of less than 5,000 Pak

Rupees) – this compares with the national average of 7,685 Pak Rupees (95). Khan et al in a study of TB knowledge awareness in 170 hospital attendees in Karachi reported that just under half (46.5%) of their patients were classified as having a low income

(77). Mushtaq et al compared TB in urban and rural communities of Punjab province and found that more than half (59.3%) of rural community cases had a low income (75).

This means that the majority of patients in rural areas of Pakistan are classified as having low income. Since 75% of KPK population live in rural areas, low income level could be one of the risk factors for this higher notification of TB in females in KPK.

4.1.5 Ethnic Origin

The majority of patients with TB in the reviewed studies were Pakistani – Shah et al determined that of 386 single male inmates 92.5% (n = 357) were Pakistani and 7.5

% (n = 29) were foreign nationals: unfortunately at the national level, NTCP do not report the TB incidence of foreign nationals living in Pakistan (54). There are about 0.6 million registered Afghan refugees living in KPK, but they have a separate parallel programme for TB diagnosis and treatment. Therefore, we can assume that ethnic origin has no association with higher female notification of TB in KPK.

4.1.6 Summary

The papers reviewed in this section have suggested that male rates for reported

TB notification in Pakistan are higher than KPK female rates. Research regarding gender and TB in KPK is very limited. However, available published studies and data from National TB Control Programme have described the higher notification rates of

TB in women, supporting the premise of this thesis that reasons behind the variation in

TB notification in parts of Pakistan require further investigation. Socio-economic factors are likely to explain the differential notification rates that we wish to investigate. 78

4.2 Risk Factors for Disease Acquisition

4.2.1 Introduction

TB is a major cause of death accounting for 7% of all causes of mortality (99).

NTCP data suggests that overall males were more likely to be affected than females in

Pakistan. However, there were several studies that pointed to a higher case notification of females than males in the province of KPK (6)(92)(96). In this section I will discuss the main risk factors for TB disease acquisition in Pakistan. The majority of females in

Pakistan are unemployed and living as housewives in joint family system (where entire family of brothers, sisters and parents live together in a single house even after marriages and children) (100). Therefore, I have assumed that household contact, overcrowding, malnutrition, smoking and male dominant culture are potential risk factors for Tuberculosis in Khyber Pakhtunkhwa.

4.2.2 Disease Risk Factors

4.2.2.1 Household Contact

Akhtar et al explored the role of household contact as a risk factor for acquiring TB in the Umerkote distract of Sindh. Out of 385 study subjects, more than half (56.0%) reported living in the same bedroom as someone with TB. The authors reported that nearly half (n=184, 47.8%) of all children with TB were most likely infected as the result of a contact in the family (the mother or father) having active disease. More females (8.8%) than males (7.3%) also claimed they had acquired their

TB from their married partner (68).

In a cross sectional study of TB patients seen at the Hyderabad DHQ hospital between August and November 2008, Soomro and Qazi determined that 26 (68.4%)

79 female cases reported household contact with a TB patient as a possible source for their infection (95).

Shah et al identified 15 cases of TB in Karachi out of 386 single male inmates

– in this case control study a family history of TB was significantly more likely to be associated with confirmed disease (adjusted OR 7.78, 95% CI = 1.12 – 62.48) (54).

Recently, Aziz et al in study of 200 people selected from the family members of TB patients identified household contact as a major risk factor for developing TB (as determined by a positive Mantoux test); the authors suggested that a household contact, where there was an index case with active disease, was nearly 13 times more likely

(p<0.01) to have a positive Mantoux test than controls drawn from households without such a case (70). The majority of females are unemployed, spending most of the time inside the home. This suggests that if there are undiagnosed or female patients in the home with active TB, the risk for TB occurrence will increases in females. Therefore, I consider that the time spent by patients outside of the home is one of the variables for my study.

4.2.2.2 Overcrowding

When more people are living in a single house where limited space is available to them, as compared to a WHO standard, this is referred to as overcrowding. The WHO recommends a floor space area with 11 or more square metres for two persons, 9 to 10 or more square metres for 1.5 persons and 7 to 9 or more square metres for one person, while children 1 to 10 years counted 0.5 unit and babies under 12 months are not included (101-102). Several studies have explored the role of overcrowding as a risk factor for Tuberculosis. In a cross sectional study conducted by Soomro & Qazi in the

Department of Medicine in DHQ hospital Hyderabad between August and November

2008 on TB patients, all the participants of the study reported living in overcrowded

80 housing, (defined as more than two persons of different genders sleeping in the same room) (95). Similar findings were reported by Javid et al in a cross sectional prevalence study of drug resistance in NWFP and suggested that almost all (n=110, 92.4%) the culture positive patients were from overcrowded families (71). Shah et al studied the transmission of Tuberculosis in 386 prisoners in Juvenile prisons in Karachi and reported that overcrowding was the main risk factor for disease acquisition (54). From the above evidence, it therefore concluded that overcrowding is also one of the most important risk factors for Tuberculosis and may be responsible for higher notification of

TB in female patients in KPK.

4.2.2.3 Smoking

Smoking is one of the main risk factors of Tuberculosis. Smoking not only increases the risk for TB in smokers but children and other family members exposed to passive smoking are also at a risk of TB (103). Three studies have looked at the role of smoking and occurrence of TB. Soomro & Qazi reported that more than half (58.0%, n

= 38) of males with TB reported that they smoked – this was considerably higher than the corresponding figure for females (21.1%, n = 8, p= 0.01) (95).

Hussain et al studied latent TB in prisoners; they randomly recruited 425 participants from five main provincial prisons. One hundred and sixty four (38.6%) were from Abbottabad, 19.5% (n = 83) from Haripur, 21.1% (n = 90) from D.I Khan,

12.7% (n = 54) from Kohat and 8.0% (n = 34) were from Mardan. The authors reported that more than half (53.0%) of the participants were smokers with 29.0% (n = 77) saying that they smoked between 1 and 5 cigarettes daily (72). In another study (Shah et al 2003) in Karachi showed that 42.6% (n = 164) of the participants with suspected TB were smokers; these results are similar to the findings of Khan et al who reported that

40.0% males and 8.0% females in Pakistan smoke regularly (54)(77).

4.2.2.4 Malnutrition

Malnutrition is a condition caused by lack of proper nutrition or not having enough food to eat. Malnutrition was identified as a major risk factor for TB by Soomro and Qazi in a cross sectional research study focusing on gender differences in relapse patients. One hundred relapse cases were included in the study, with more males (38 out of 62, 61.3%) than females (18 out of 38, 47.4%) reporting malnutrition as a risk factor for disease acquisition (95). Research conducted in India reported that malnutrition is not only a risk factor for TB but better diet also plays an important role in the treatment of TB (104). Malnutrition is mainly associated with poverty and it is evident from the literature review that the majority of TB patients in Pakistan belong to a low income group. Therefore, malnutrition in my study will also be assessed by asking questions about eating meat on daily, weekly and monthly basis.

4.2.3 Cultural Risk Factors

4.2.3.1 Male Dominant Culture

Ullah et al in a prospective study carried out in LRH (Lady Reading Hospital) on 525 patients, found that 66.5% (n = 349) female and 33.5% (n = 176) males were diagnosed with Extra Pulmonary Tuberculosis (EPTB). The authors suggested that the higher incidence of TB in the region was due to the male dominant culture, where males have a higher priority than females – thus it is common practice in Pakhtuns’ families – for example, males are served food earlier than females, and male children have a higher standard of health, education and emotional care as compared to females (51). In a local study of urban and rural differences in KPK, Rahman et al also reported that male gender is dominant in making decisions and in the social life (105).

4.2.3.2 Summary

Articles reviewed in the literature review showed that overcrowding, malnutrition, household contact and cultural restraints were main risk factors for TB in

KPK.

4.2.4 Knowledge Based Risk Factors

4.2.4.1 Patient's Knowledge about TB

4.2.4.1.1 Introduction

This section will consider patient awareness and knowledge about TB – particularly the signs and symptoms of TB, transmission and prevention of TB, stigmatisation about TB and urban and rural differences in disease occurrence.

4.2.4.1.2 Signs and symptoms of TB

In a very recent study, Mushtaq et al compared the knowledge, attitude and training about TB in patients: a total of 1080 (urban 66.2%, rural 63.4%) respondents above 20 years were recruited randomly from two districts of Punjab Province. Just under two thirds of all respondents reported that cough was the main symptom of pulmonary tuberculosis, with significantly fewer in each group (53.5% urban and

36.1% rural, p=0.001, 95% CI, 1.59 – 2.61) correctly identifying that a cough lasting three or more weeks was the more accurate indication of pulmonary TB (75). Lack of knowledge about signs and symptoms of TB may lead to delay in the diagnosis and treatment. This is potentially therefore a very important study because it included both rural and urban populations but the research was limited to only two districts of Punjab province.

In a hospital based cross sectional study, Samera et al looked at the relationship between knowledge of TB and time to diagnosis and treatment. Three hundred and one patients presenting to the outpatient departments for chest diseases at Nishter Hospital

Multan and Bethania Hospital Sialokot were recruited to the study. The researchers found that of those who presented within 20 days of onset of illness, 63.0% were aware that cough was a principal feature of TB, compared with only 37.0% in those patients who presented after 20 days – these observed differences were statistically different, p=

0.007 (76).

Khan et al studied knowledge of TB among 170 TB patients at out-patients department (OPD), attending two different hospitals in Karachi. The percentage of patients able to identify the signs and symptoms of disease were as follows: cough

(83.5%), fever (54.7%), chest pain (24.7%) and blood in sputum (24.7%) (77). The study was limited to Aga Khan and Jinnah hospitals in Karachi. Aga Khan is a private teaching hospital and the majority of people coming for treatment belong to a high income group. It is highly likely that the sample of patients in this study is not representative of the overall population of Karachi.

4.2.4.1.3 Urban and Rural differences

Mushtaq et al in cross sectional study carried out in two districts, explored the knowledge and behaviour patterns about TB in urban and rural communities of Punjab province. Due to a generally higher level of education, the urban population were more aware about TB. Two out of 35 distracts in Punjab were randomly selected. Five Union

Councils, two from urban and three from rural areas were randomly selected for this study. From each Tehsil (sub district), two villages were selected for interviewing 18 individuals, yielding a total of 1080 participants. The majority (n=648, 60.0%) of the

84 study subjects were recruited from rural as compared to urban areas (n=432, 40.0%)

(75).

A study of 754 female subjects reported that 76.5% of urban dwellers compared with 57.7% of rural dwellers correctly identified that cough (of any duration) was the main symptom of TB. Furthermore, urban women were more informed about

TB than rural women with 47.8% reporting that the lungs were involved in TB infection compared with 13.3% of rural females (p<0.05) (79). In a local study, results showed that people living in rural areas were significantly (p = <0.05) illiterate as compared to urban population in KPK (105).

4.2.4.1.4 Transmission and prevention of TB

Agboatwalla et al conducted research on gender and urban/ rural differences in knowledge about TB. In a questionnaire-based study, 51.0% of the 754 participants were male. The authors reported that just over three quarters of females from rural areas were classified as illiterate compared with 51.0% of urban-based females: in addition, female respondents from rural areas (24.5%) were less likely to report that TB was spread from TB patients through coughing and sneezing than urban living females

(38.9%) - these observed differences were statistically significant (p= 0.004) (79).

In another cross sectional study of knowledge about TB transmission in two districts of Punjab Pakistan, just over four out of five (80.8%) of urban participants randomly selected from the community, correctly identified that TB was spread by the respiratory route (droplet nuclei generated by coughing and sneezing) compared with

68.5% of rural participants – this observed difference was statistically significant

(p<0.001) (75).

Similarly, results explored knowledge about TB in patients and their relatives: more males (n=131, 65.0%) were interviewed than females (n=72, 35.0%).They 85 reported that more than four out of five participants (n=166, 82.0%) were aware that TB was a communicable disease and that just over half (n=110, 54.0%) said that it was spread through respiratory droplets (78).

4.2.4.1.5 Stigmatisation about TB

Results from the research conducted in two districts of Punjab (Nankana Sahib and Bahawalnagar) showed that stigmatisation (shame or disgrace associated with TB) was a common cause of delay in the diagnosis and treatment of both urban (83.6%) and rural (71.5%) patients – this observed difference was statistically significant (p<0.001)

(75). Samera et al also identified stigma in 40.0 % of the educated population and suggested that patients due to shame and fear of being rejected by the community do not inform even their family members about TB. This can result in delay in diagnosis and treatment (76). These results are similar to the research conducted by Khan et al in

Karachi who reported that just over two fifths (n=70, 41.0%) of the 170 in-patients studied did not tell relatives they had TB because of the stigma associated with the disease (77).

4.2.4.1.6 Summary

The literature on patients’ knowledge about Tuberculosis is very limited. In

Khyber Pakhtunkhwa, I could not identify any research on the knowledge of TB patients. However, there were several studies carried out on patients’ knowledge in

Pakistan. It is evident from the results of studies included in the literature review that patients’ knowledge regarding TB is very limited, especially in rural communities.

Since nearly 75 % of KPK population live in rural areas, it is a reasonable assumption that patients from rural districts of KPK have poor knowledge of TB and should be investigated for the association between knowledge and higher notification of TB in females. 86

4.2.4.2 Physicians’ Knowledge about the disease

4.2.4.2.1 Introduction

In this section I will review the studies that have explored General

Practitioners’ awareness of the signs and symptoms of TB, approaches to diagnosis, and the appropriate treatment and follow-up of patients – the physicians studied were drawn from either Government or private hospitals.

4.2.4.2.2 Signs and Symptoms of TB

Ahmed et al looked at the knowledge of male Private Practitioners (PPs) providing health care services in the District of Thatta, a rural area of Sindh. Of the 100

PPs who took part in the study, half (50.0%) reported they regularly treated patients with TB: 22 private practitioners (PPs) were included in a more detailed study. The authors reported that 68.0% of respondents correctly identified that a cough for ≥ 3 weeks was the main symptom of TB, with 32.0% saying that cough, together with haemoptysis, weight loss and an elevated temperature – particularly in the evening – were additional disease features (80).

Shah conducted research to assess the level of knowledge and practices among

PPs in the two cities of Punjab, Lahore and Rawalpindi. Two hundred and forty five PPs were randomly selected – 80.0% (n=197) were males: 59.2 % (n= 145) were from

Lahore and the remainder from Rawalpindi. Of the total, 17.0% of PPs in Rawalpindi and 14.0% in Lahore were reported to have a postgraduate qualification. Fifty eight per cent of PPs and 41.0% from Rawalpindi and Lahore respectively reported that they had provided TB services for between 1 and 5 patients in the last quarter before the study.

Assessing the knowledge regarding TB, only one PP reported having a cough above 3 weeks as a major symptom of the disease, while 45.0% said a cough with haemoptysis were the main sign and symptom of TB. The authors commented on the small number 87 of postgraduates who were able to describe a cough lasting for longer than 3 weeks as a major symptom of TB (83). Studies above were limited to the assessment of private practitioners’ knowledge and conducted in single district. Similarly, Ahmed et al evaluated the knowledge of male PPs only.

Khan et al explored the knowledge and practices of PPs around the diagnosis and treatment of Tuberculosis using a questionnaire-based methodology and conducted from May to September 2002 in Karachi. A total of 141 registered PPs were identified within a two km circle of the central private health facility from where the research study was carried out. All the participants were asked to join a workshop about

Tuberculosis management and requested to complete the study questionnaire: only

85.1% (n=120) PPs participated in the study by completing the questionnaire. The researchers reported that nearly all participants (95.0%, n=114) knew that cough was a major symptom of TB, whilst 86.7 % (n=104) also said that fever and weight loss (74.2

%, n= 89) were features of the disease (84).

In another cross sectional study about the knowledge and practices of 150 private practitioners treating between seven and ten TB patients monthly in Karachi, the authors reported that more than half (55.0%) of PPs said they suspected tuberculosis on the basis of respiratory symptoms including a prolonged cough, haemoptysis, productive sputum and chest pain (86).

4.2.4.2.3 Diagnosis of TB

Research conducted on 100 private practitioners in the Thatta District, a rural area of Sindh, found that the majority (86.0%, n=19) advised that diagnosis of TB required a combination of tests including a chest X-ray, erythrocyte sedimentation rate

(ESR) and complete blood count (CBC), but only 14.0% (n=3) correctly identified the need for sputum microscopy (80).

Shah et al reported that 80.0% of PPs made the diagnosis of TB and started treatment on the basis of a clinical history, Chest X-Ray, ESR and CBC (termed “self diagnosis and management”), without referral to a TB centre although the latter are responsible for treatment at the district level. Eleven per cent said they made the diagnosis of TB as above but referred the patients to the district TB centre for confirmation and treatment, and 8.0% sent the patient directly to the TB centre. Very few PPs (2.0%, n=5) had their own diagnostic facilities, and only two PPs were able to examine sputum for Acid Fast Bacilli. Only one out of 245 PPs advised sputum microscopy alone for the diagnosis of TB, while 45.0% mentioned combination of

Chest X-Ray, ESR and sputum examination (83). In another study Khan noted that just over half (55.0%) of PPs used a “self diagnosis and management” approach to TB with

21.0% saying they referred patients to the TB centre. More than half 58.3 % (n=70) PPs reported they diagnosed TB on the basis of sputum examination and 20.0% (n=24) by

X-ray chest (84). Similar observations were made by Rizvi and Hussain, who studied the diagnostic skills of 150 General Practitioners (GPs) treating on average 7 to 10 TB patients monthly in Karachi. They noted that for the diagnosis of TB only 21.0% PPs suggested sputum examination and 15.0% a Chest X-Ray (86). Hussain et al carried out a cross-sectional study to assess the diagnostic capability of PPs. The authors reported that only one out of fifty three PPs advised on the need for sputum examination for the diagnosis of TB, with liver function tests and an ESR being advised by 5.7% and 11.3

% of PPs respectively (82).

Shehzadi et al using a cross sectional research study design investigated the knowledge and practices of GPs in NWFP and Northern areas in 2003. Eighty eight study subjects were recruited by convenience sampling from Gilgit, Hunza, Abbottabad,

Haripur and Peshawar. The majority (87.5%, n=77) reported having attended chest X-

Ray interpretation classes during their medical training. More than three quarters

(77.3%) stated that their approach to management was based on “self diagnosis and management” without referral to a TB centre. The authors also reported that only 43.3%

(n=38) identified sputum examination and 21.6% (n=19) Chest X-Ray as the two principal investigation tools of choice for the diagnosis of TB (81).

It was suggested previously that patients living in rural areas belong to low income groups. WHO recommends sputum microscopy for the diagnosis of TB.

Unnecessary diagnostic tests may increase the financial burden on poor TB patients and delays the diagnosis and treatment as well (89).

4.2.4.2.4 Treatment of TB Patients

Research conducted on 100 PPs in the Thatta District, a rural area of Sindh, found that just under three fifths prescribed the correct treatment regimen for the continuation and 73.0% for the intensive phase. TB treatment takes six months where the first two months is called intensive phase and the rest of the four months is known as the continuation phase (80).

In another study of assessment of GP's knowledge in NWFP and Northern areas about the management of TB patients, when asked to write a prescription for a 60 kg new adult TB patient, only 35.2% (n=31) prescribed an accurate combination and dosage of drugs for the intensive phase and 29.5 % (n= 26) for the second phase of the treatment The authors reported that just over two-thirds (68.8%, n=60) said that the same treatment regimen should be given to pregnant women, with 10.2% (n=9) advising a termination of pregnancy, and 8.0% (n=7) suggesting that treatment should be discontinued (81). Even fewer PPs (3.7%, n=2) were able to write a correct prescription as per NTP guidelines in a cross sectional study carried out by Hussain et al. Just under three quarters (70.0%) PPs prescribed medicine additional to TB treatment – for example Ciprofloxacin (82).

The majority of PPs working in Lahore and Rawalpindi also failed to follow the NTP guidelines for TB treatment Shah et al reported that 68.0% advised a combination of four drugs, while 15.0% chose three separate drugs for TB treatment

(83). It is surprising that 87.0% of the GPs did not differentiate management for initial and continuation phases, while only 6.0% mentioned a two month required period for the first phase of TB treatment.

Khan et al explored the prescribing practices of PPs in the Aga Khan

University hospital in Karachi. They reported that the majority (79.0%) of PPs prescribed four drugs for the initial phase of TB treatment as per NTP guidelines although just under a half prescribed less than the recommended dose of Pyrazinamide

(43.0%) and Ethambutol (48.0%) (84). Similar findings were reported in a questionnaire based research conducted from May to September 2002 in Karachi by

Shah, who suggested that just over a quarter (26.7%) of PPs did not follow the recommended regimen of four drugs in the first phase of TB treatment with 35.0 %

(n=42) also prescribing an incorrect regimen for the second phase. The study also found that of the 102 participants, 20.0% prescribed a lower than recommended dose of

Rifampicin with 49.0% prescribing for longer than the recommended duration (83).

A study of practice assessment of 150 GPs treating 7 to 10 TB patients monthly in Karachi, reported that only 23.0 % (n=34) prescribed the correct treatment regimen for TB with 29.0% of GP failing to answer this question. This highlights deficiencies in the treatment of TB by GPs. Replying to the duration of treatment, 24.0% reported 12 months and above which is alarming because of the potential for causing MDR-TB

(86).

Poor diagnosis and treatment knowledge are therefore important factors responsible for the delay in the diagnosis and treatment of TB patients.

4.2.4.2.5 Directly Observed Treatment Short Course

Directly observed treatment short course (DOTS) has been recommended as a treatment strategy for TB since 1993, when TB was affirmed as a global emergency.

Several studies have explored the knowledge and practices of medical practitioners regarding TB DOTS in Pakistan. In District Thatta province of Sindh (rural area), researchers explored the knowledge of 100 PPs in relation to TB DOTS. Half (50.0%) of the PPs reported treating TB patients but no physicians identified their practice as following recommended DOTS strategy. Furthermore, few knew how to trace defaulters, because they never maintained TB patients’ records (80). Shah et al found similar findings in a research project to assess the level of knowledge and practices of

PPs in two major cities of Punjab. Out of 245 PPs, nobody advised treatment under

DOTS (83).

In 2005, Pakistan achieved 100% DOTS coverage nationally (1). Further, due to the lady health worker (LHW) programme the implantation of DOTS at village level become more effective. However, my own observations suggest that the use of LHW in rural and remote areas of Pakistan still needs improvement

4.2.4.2.6 Follow up of TB Patients

In a rural area of Sindh (District Thatta), Ahmed et al (2009) explored the knowledge of 100 PPs about TB patient follow-up. Half (50.0%) reported treating TB patients, but only 40.0% followed the patients for the duration of treatment. The majority (n=19, 86.0%) never advised sputum examination at end of the second, fifth and seventh months (80). Similar findings were reported by Shah et al who conducted research to assess the level of knowledge and practices among PPs in Lahore and

Rawalpindi. Slightly more than three quarters (76.0%) preferred a chest X Ray for the assessment of TB, whereas WHO only recommend sputum examination at the end of

92 second, fifth and seventh months (83). Khan et al also identified that only 35.0%

(42/120) of doctors were aware that sputum examination is the most useful test in follow up (84).

National TB control programme recommends follow up at the end of the second, fourth and sixth month for sputum negative and fifth and seventh month for sputum positive patients (106). It is one of the most important components of TB DOTS programme and negligence in follow up may lead to multi drug resistance TB (MDR-

TB).

It is evident from the above studies that physicians’ knowledge is limited about diagnosis and treatment of TB. However, the majority of the respondents in the studies reviewed were private practitioners. Research conducted in seven countries including

Pakistan, showed that delay in the diagnosis and treatment was significantly higher in private clinics (97). TB patients in Pakistan prefer public facilities (TDCs, BHUs, RHCs and Hospitals) for free diagnosis and treatment of TB. National TB control programmes have TB diagnostic and treatment centres (TBDCs) for the diagnosis and treatment of

TB patients in KPK. Therefore, it is important to evaluate the knowledge of physicians working in those TBDCs about the diagnosis and treatment of TB.

4.3 Summary of the Literature Review

Chapter 4 has explored the literature in relation to the epidemiology, and risk factors for tuberculosis in the Pakistan (See Table 13 and Table 14 on pages 63-68).

Patient's knowledge and treatment seeking behaviour of TB was also discussed in this section. (See Table 15 and Table 16 on pages 70 and 71)

Finally the awareness and practices of private and public practitioners regarding TB management was explored in detail and the results identified a number of serious gaps in knowledge and practices 93

CHAPTER 5: METHODOLOGY

5.1 Introduction

This chapter will describe and explain the research methodology, aims and objectives of the study. Research questions and methods for each objective will be discussed. The process of data collection, data collection instrument, ethical consideration and data entry will also be discussed.

5.2 Aim of the Study

To explore the reasons behind the variation of Tuberculosis Notification between males and females in the province of Khyber Pakhtunkhwa (KPK), Pakistan.

5.3 Objectives of the Study

1. To describe the epidemiological characteristics of confirmed TB Cases in

Khyber Pakhtunkhwa by age, gender, nationality, geographical location and

disease type for 2002 to 2010.

2. To explore risk factors for newly diagnosed laboratory confirmed TB disease

in all adult patients in KPK from July 01, 2012 to September 30, 2012.

3. To estimate the level of knowledge about TB in adult TB patients in KPK from

July 01, 2012 to September 30, 2012.

4. To evaluate the level of knowledge, attitude and practice of Physicians working

in TB diagnostic centres (TDCs) in KPK.

5.4 Research Questions

Keeping in view the aims and objectives, the issues and questions were discussed in detail with the supervisory team and I settled on the following four research questions around which I shaped the study design: 94

1. What are the Epidemiological Characteristics of Confirmed TB cases in KPK?

2. What are the risk factors for TB in adult patients in higher versus lower

notification districts within KPK?

3. What do adult TB Patients in KPK know about their disease?

4. What are the levels of knowledge, attitude and practice about TB for

Physicians working in TBDCs in KPK?

5.5 Hypothesis

On the basis of these research questions the following four hypotheses were developed:

1. The notification of TB in female patients is higher than males in the province

of KPK.

2. Socio-economic issues are the most important risk factors for higher

notification of TB in females in the province of KPK, Pakistan.

3. Females of KPK are less knowledgeable than males regarding Tuberculosis.

4. General practitioners’ knowledge of TB in respect of diagnosis, treatment and

monitoring is not satisfactory in the province of KPK.

5.6 Methodology for Objective One

5.6.1 Study Setting and Target Population

KPK is one of the provinces in Pakistan and is divided into 26 districts for administrative purposes. According to the 1998 census, the population of KPK is 17.7 million (29). The executive district officer health (EDOH) is the person who is in charge of health-related projects in the entire district. The district TB control officer (DTCO) is responsible for the diagnosis, treatment, management and case reporting activates at district level for National TB Control Programme (NTCP). There are 225 TB diagnostic 95 centres (TBDC) in KPK (107-108). Patients who present to all 225 TB diagnostic centres in KPK for diagnosis and treatment were reported to NTCP on monthly, quarterly and annual basis by DTCO. The study population included all registered patients who reported for treatment to NTCP from 2002 to 2010 in KPK.

5.6.2 Study Design

Through a retrospective study, I performed a detailed preliminary quantitative analysis of existing NTCP confirmed TB admission records to identify trends in TB outcomes according to age, gender, location, nationality, disease type, diagnostic tests and the post treatment cure rate using NTCP data from 2002 to 2010. The data helped me in identifying epidemiological characteristics for confirmed TB cases in Khyber

Pakhtunkhwa.

5.6.3 Data Collection

The National TB Control programme (KPK) collected data from all districts of

Khyber Pakhtunkhwa on regular basis through the DTCO. Specific WHO TB data collection instruments have been used for data collection (Copies attached as Appendix-

14.7 on page 226). Monthly data received from district TB control officer between 2001 and 2009 were entered by NTCP data operators into a Microsoft Access© database especially designed for NTCP KPK. In 2009 an electronic reporting system (ERS) was developed for recording data of TB patients in KPK.

Data related to Objective One was obtained through analysing the NTCP

Microsoft Access© database data for TB patients from 2002 to 2009. Data for 2009 to

2010 was retrieved from the ERS database of NTP. Access to the database was provided by the NTCP IT department with a valid user name and password. The data was

96 analysed and tabulated to show the distribution of TB by age, gender, type of disease, geographical area, cure rate, default rate, death rate.

5.6.4 Data Quality

The data presented was checked for accuracy and entirety. To identify any deficiencies in the data, I visited the NTP Office and district TB control offices of five districts - Abbottabad, Bannu, Dir (lower), Kohistan and Peshawar. TB registers of the selected districts were then manually checked to verify the NTP data for confirmation.

5.6.5 Analysis

Study Data from 2002 to 2010 were exported to Microsoft Excel 2007 which was then used to aggregate and tabulate the results. Patients’ age, sex, region, type of disease and treatment outcome were tabulated. Statistical analyses were carried out using SPSS 22.

5.6.6 Ethics

The ethics committee of the Provincial Tuberculosis Control Programme

Khyber Pakhtunkhwa approved this study and agreed for the additional support with regard to field work in 2013. Authorisation was acquired to conduct the study and communicated with the EDOs (H) of all districts oh KPK. (Copy of the NTCP permission letter is attached as appendix 14.2 on page 196).

5.7 Methodology for Objectives Two and Three (Study Design)

5.7.1 Study Setting and Target Population

For the purpose of data collection ten districts were selected, five from higher and five from lower notification areas. (See Table 17 below) These districts were identified as higher and lower TB case notification regions through the earlier descriptive study using the NTCP data from 2002 to 2010, in respect of gender difference of TB in KPK. Five TDCs were further selected randomly from each district for the recruitment of study subjects. All confirmed adult TB patients presented to each

TDC between 1 July 2012 and 30th September 2012, were included in the study.

Table 17: Five Higher and Five Lower Notification Districts of KPK

Higher Case Notification Districts Lower Case Notification Districts

Kohistan Karak Battagram Kohat Chitral Mardan Buner Nawshera Upper Dir Malakand

Key to text colour in table:

 = Districts with Lower Female TB Notification

 = Districts with Higher Female TB Notification

5.7.2 Study Design

A case control study was chosen to explore the reasons behind the variation in case notification of TB patients in KPK. Participants registered to higher case notification districts (Kohistan, Battagram, Chitral, Buner and Upper Dir) were marked as case group and patients reported to lower case notification districts (Karak, Kohat,

Mardan, Nawshera and Malakand) were noted as control group for the study.

5.7.3 Sample Size

The results from objective one showed that a total of 221,051 TB patients were reported in KPK between 2001 and 2010. Therefore, dividing this figure by nine (years) would suggest that on average 24,561 TB patients notified annually and 2,047 monthly in KPK during 2001 to 2010. As there are 225 TDCs in KPK, by dividing this monthly figure by the total number of TDCs, on average nine patients reported to each TDC monthly. By excluding cases below 15 years of age, relapsed patients (i.e. those who had taken full course of TB treatment in the past and were declared cured or treatment completed but re-infected), transferred in (had taken TB drugs and had been transferred from another TDC), treatment failure (smear positive patients who had taken TB drugs for five months or more) and treatment default (had taken TB drugs for a certain period then interrupted for two or more months) cases I expected to receive five patients monthly to each TDC. As in each district there were five TDCs, the anticipated sample size was calculated as 375 (five patients from each TDC monthly multiplied by twenty five TDCS in each group and multiplied by three months) new adult TB patients from higher (Case Group) and 375 from lower notification regions (Control Group) from 1st

July 2012 to 30th September 2012. All newly diagnosed TB patients confirmed by sputum microscopy in the selected 50 TDCs from ten districts were recruited between

1st July and 30th September 2012.

5.7.3.1 Inclusion Criteria

 All adult newly diagnosed male and female TB patients confirmed by sputum

microscopy, because I believed that defaulted or previously treated patients

would be more knowledgeable than newly diagnosed ones due to their lived

experience of treatment. (‘Defaulted’ are those cases who do not complete their

treatment), (‘previously treated’ are those treated but not cured or who are re-

infected).

 All adult patients permanently living in the TDC area of the district. I include

those patients because it was possible for community health workers (CHWs) to

trace and meet the patients at their home for confirmation of data.

 Willingness of subjects to take part in the research, because patients have a right

to accept or reject the interview with the assurance that their right to receive

health care will not be affected by the decision.

5.7.3.2 Exclusion Criteria

 Patients below the age of 15 years were excluded from the study.

 Referred or transferred patients from other TDCs or districts (Those who were

diagnosed in other centres or other districts).

5.7.4 Date Collection Tools

Structured questionnaires were used for data collection using face to face interviews with patients attending the TDC. A questionnaire was used because it made data capture relatively straightforward. The questionnaire was developed in Epi Info 7© software developed by the Centre for Disease Control in the USA (109). The advantage of using Epi Info© was that, it automatically created a database for analysis of inputted data. A further consideration in using a questionnaire was because research by

Saunders, Lewis and Thornhill (110) suggested that if study subjects are illiterate, planned and then administered questionnaires are the best choice for information collection in research.

The research tool was divided into four sections:

100

5.7.4.1 Baseline Data

Variables included in this section provided information about case no, date of registration, name, gender, age, height, weight, type of TB and district of residence.

Age, height and weight were used for body mass index (BMI) calculation.

5.7.4.2 Demographic Data

This section contained demographic information of patients, such as educational status, marital status, economic status, number of family members, number of rooms, type of fuel of using for cooking, eating meat interval etc.

5.7.4.3 Risk Factors

Questions explored past medical history and some potential risk factors. These variables were mainly past and positive family history for TB, associated diseases, smoking etc.

5.7.4.4 Knowledge of TB

To assess the level knowledge of TB patients, questions were included to find data about the cause, transmission and spread of TB. Questions were also asked about the most common symptom of TB, duration, cost, and consequences of incomplete treatment (copy of questionnaires attached as Appendix 14.4 on page 205).

101

5.7.5 Validity of Data Collection Tool

The degree to which a data collection tool measures what is supposed to be measured is called validity. The spectrum of validity is mainly dependent on three points: the structure of the questionnaire, the purpose of the questionnaire and the population which will be investigated. For the reliability, I validated the questionnaire using the following main methods:

5.7.5.1 Content Validity

After selecting different variables, based on the literature review, a structured data collection tool was developed. Initially, content validity was achieved through detailed discussion with PhD students (a primary expert group) who were working in the same research field. This had the benefit of being able to discuss the tool with other researchers and ‘test it out’ prior to the formal pilot (111).The questionnaire was then discussed with a secondary expert group (the supervisory team) for further content validity.

5.7.5.2 Concurrent Validity

The structure of my questionnaire was mainly based on the variables extracted from the literature review. Therefore, it could be argued that the validity of the variables was already established. However, due to the addition of extra questions proposed by expert groups, further validity testing of the questionnaire was advised by supervisory team.

5.7.5.3 Piloting the Questionnaire

For final validation of the data tool, I tested the questionnaire on 50 medical technicians recruited from selected TDCs of ten districts in KPK before and after the training sessions in each district. All medical technicians were trained and the 102 questionnaire was discussed with them in detail. In addition, after completion of training sessions, 5 patients were randomly selected from each district (total of 50) and medical technicians were asked to conduct face to face interviews to test the questionnaire practically. The purpose of this activity was mainly to achieve validity and:

 To pilot test the questionnaires for data collection.

 To estimate the time duration for data collection.

 To estimate the level of difficulty for words or terms used in questionnaires

for medical technicians.

 To increase their understanding of the field work.

 To pilot test the questionnaires for data collection in respect of patients’

point of view.

 To develop the questionnaire for final approval.

5.7.6 Data Collection

5.7.6.1 Face to face interview

Since the majority of the population of KPK are illiterate, therefore face to face interview was the chosen method for data collection. Before interview, all patients were informed in detail about the research activity and verbal consent was obtained. All confirmed new TB patients were interviewed by trained medical technicians in selected

TDCs between 1st July 2012 and 30th September 2012. To provide flexibility and confidentiality, medical technicians ensured that all interviews were conducted in a safe and private place. All female patients were interviewed in the presence of LHWs.

103

5.7.6.2 Double Data Entry Strategy

Area supervisors were trained to use Epi Info 7© for data entry on a weekly basis. To avoid and minimise human error during the data entry process, a double data entry strategy was adopted. Both male and female area supervisors entered the same data separately on an identical Epi Info-7© database. At the same time, I also entered data separately on a weekly basis using the same Epi Info 7© database. As a result of this, two data sets were created. After completion, 55 different discrepancies were found, which were reconciled by confirming against the original questionnaires.

5.7.6.3 Analysis

Data was collected through pre-coded questionnaire and analysed using Epi

Info 7© software. Chi-square tests were used for comparisons between two groups.

Multiple logistic regression was used to test for the simultaneous involvement of several factors that might influence the higher notification of tuberculosis in females in the province of Khyber Pakhtunkhwa, Pakistan. A P value < 0.05 was considered statistically significant. For the determination of relative risks (RRs), confidence intervals and standard errors, Stata© software was used.

5.7.7 Ethical Approval

5.7.7.1 Ethics Committees

We sought and received approval for conducting this research from the

National Tuberculosis Control (NTCP) Khyber Pakhtunkhwa. (A copy of approval is attached as Appendix 14.2 on page 196).

We also obtained ethical approval from the Committee on Ethics of Research on Human Beings at The University of Manchester to conduct the study (See Appendix

14.1 on page 194 for the copy of Ethical approval). 104

5.7.7.2 Informed Consent

Obtaining consent from the patients was problematic because of the high levels of illiteracy. We agreed with the ethics committee that patients will be verbally consented and that the medical technicians and LHWs were specifically trained to ensure that patients were not coerced into taking part.

We did not anticipate any problems in relation to obtaining the information because the questionnaire was primarily asking about knowledge and socio-economic factors, much of which was already collected as part of the health surveillance carried out by the NTCP (See appendix 14.4 on page 205).

5.7.7.3 Confidentiality

NTCP already follow the guidelines set by the World Health Organisation

(WHO) regarding ethics and health safety within their premises. Staff who were carrying out this research were already subjected to the requirements for confidentiality regarding patient data, public safety and ethics. Additional training was provided to all the research staff specific to our research project with specific emphasis being given to consent and confidentiality

5.8 Methodology for Objective Four

5.8.1 Research Plan

One of the reasons that more women were diagnosed with TB could be related to the diagnosis and treatment carried out by physicians working in the primary health care setting. Understanding physicians’ knowledge, attitude and practice will be critical to understanding why women have a higher notification of TB. We therefore carried out a cross sectional study targeting the general practitioners of selected TDCs. Delay in diagnosis and treatment increases the chance of spreading infection to healthy family 105 members. As in KPK, the majority of females are not working and they remain at home.

This suggests that a case of active TB in an untreated woman may increase the chance of further TB infections in other women.

5.8.2 Study Setting and Target Population:

TDCs that were previously selected for data collection in objective two and three were included in this part of the study. All general practitioners working in the

TDCs that form the sample for this study were asked to fill in a questionnaire which explored their knowledge, attitude and practice in relation to treating female patients.

This part of the study objective was completed between 1st August 2012 and 30th

September 2012 and was carried out by the principal investigator.

5.8.3 Study Design

A cross-sectional study design, using a pre-structured questionnaire was chosen to assess the knowledge of physicians working in the TDCs in KPK.

5.8.4 Sample Size for Objective Four

All practitioners working in the selected TDCs previously being used for data collection for objective two and three were included in the study. Fifty general practitioners were interviewed for Objective Four.

5.8.5 Date Collection Tools

Face to face interview was chosen for data collection using a pre-structured questionnaire.

I prepared a questionnaire with both open and closed questions. The tool was divided into two sections.

106

5.8.5.1 Demographic Data.

Variables included in this section provided demographic information of doctors such as, name, gender, age, medical school attended, qualification, duration of service etc.

5.8.5.2 Knowledge of TB

To estimate the level of knowledge and understanding of TB in general practitioners working in TDCs, several questions were asked (Questionnaire attached as

Appendix 14.4 on page 205).

5.8.6 Data Collection

A structured questionnaire was used for data collection with written consent.

Districts visits were arranged according the availability of doctors and appointments were scheduled for data collection. Data was collected by the main investigator through face to face interview.

5.8.7 Dual Data Entry

Data was entered into Epi Info 7© on daily basis. Dual data entry method was obtained to eliminate human error during data entry process.

5.8.8 Analysis

Statistical analysis was carried out using the Epi Info© software (version 7)

Chi-square test was used for statistical significance. A p-value <0.05 was measured to be statistically important.

5.8.9 Ethics

Ethical approval was also obtained for this part of the study (see page 104.)

107

5.9 Summary

Chapter 5 has described my research methodology. It includes my study aim, objectives and research questions. I have explained both the construction and methods for my four main research surveys. The structure of data collection tools, pilot study for testing those questionnaires and data collection in the field study were also described in this chapter.

108

CHAPTER 6: RESULTS - DESCRIPTIVE EPIDEMIOLOGY

6.1 Introduction

The study of disease incidence, prevalence and demographic factors (age, gender and geographic area) is called descriptive epidemiology. Differences in the occurrence and frequency of disease in respect of demographical factors may help in identifying the causes of a disease.

The descriptive epidemiology of TB patients in the province of KPK between

2002 and 2010 provides important data regarding the disease burden. Assessing this data may help in planning for the data collection phase relevant for objectives two, three and four.

6.2 TB Patients Registration from 2002 to 2010

In Khyber Pakhtunkhwa 221,051 total TB patients were registered between

2002 and 2010 in the provincial Tuberculosis control programme [Table 18]. Major differences were noted in figures and rates of TB cases reported in each district of the province. A large number of cases were reported in Peshawar (5763, 16.81%) followed by Mardan (3060, 8.93%) and Nawshera (2117, 6.18%). The lowest number of TB cases were reported in Tank (471, 1.37%) followed by Upper Dir (679, 1.98%). About an equal number of cases were reported from Hungo (2.25%), Karak (2.28%) and

Malakand (2.37%). Similar figures were notified from Bannu, Shangla, Kohat, and

Swabi, Swat, Abbottabad with 3.65%, 3.68%, 3.69% and 4.93%, 5.05%, 5.12% respectively [Table 18, , Figure 9].

Six out of twenty four districts accounted for 52% of total TB patients registered in the province of Khyber Pakhtunkhwa from 2002 to 2010: Peshawar

109

(16.81%), Mardan (8.93%), Nawsheara (6.18%), Mansehara (5.12%) and Abbottabad

(5.35%) ([Figure 8, Figure 9].

Four distinctive trends can be seen in the registration of TB patients from 2002 to 2010. There was a 69.00% increase reported in registration of new TB cases between

2002 and 2003. Subsequent to this, there was a decline noted in the annual increase in case notification between 2003 and 2005 (31.52%, 37.50%) when compared to the period from2002 to2003. Fluctuations in the annual notification percentages can be seen from 2005 to 2009 (i.e., 8.69%, 15.35%, 1.33% and 10.97%). The decrease in annual case notification during 2005 and 2009 could be possibly due to 100% DOTS coverage achieved in 2005. In 2009-2010 the annual TB patients registration decreased by -

0.71% [Figure 8, Figure 9].

110

Table 18: Annual Case Notification of TB by Districts in KPK, 2002-2010 No District 2002 2003 2004 2005 2006 2007 2008 2009 2010 Total

1 Abbottabad 0 740 1120 1191 1572 1733 1521 1908 1756 11541 2 Bannu 0 493 405 675 832 1116 1255 1281 1251 7308 3 Battagram 0 0 53 417 338 495 644 691 697 3335 4 Buner 74 446 446 548 610 704 727 718 832 5105 5 Charsadda 19 359 714 1101 1081 1369 1488 1773 1502 9406 6 Chitral 159 881 814 719 770 948 887 904 841 6923 7 DI Khan 911 482 661 1107 1100 1461 1294 1288 1422 9726 8 Hangu 0 0 0 229 225 282 469 573 772 2550 9 Haripur 0 767 829 1032 1149 1175 1147 1116 862 8077 10 Karak 0 0 0 303 471 515 625 770 782 3466 11 Kohat 0 68 360 622 770 1005 971 1131 1265 6192 12 Kohistan 0 0 15 431 425 415 520 683 720 3209 Lakki 13 0 0 125 654 582 858 897 1075 1034 5225 Marwat 14 Lower Dir 0 142 398 819 990 977 1093 1116 1124 6659 15 Malakand 0 30 310 464 498 632 610 703 813 4060 16 Mansehsra 0 437 1669 1329 1653 1617 1583 1945 1832 12065 17 Mardan 1307 1172 1977 2421 2231 2937 2735 3087 3060 20927 18 Nowshera 819 984 991 1362 1350 1546 1738 2115 2117 13022 19 Peshawar 2616 3546 3734 4312 4701 5647 5381 5917 5763 41617 20 Shangala 0 0 45 844 916 904 890 1228 1260 6087 21 Swabi 83 768 829 1036 1307 1334 1591 1636 1690 10274 22 Swat 2022 2009 1917 1987 2087 2028 1926 1729 1730 17435 23 Tank 0 0 21 262 331 381 407 458 471 2331 24 Upper Dir 0 213 372 618 624 620 709 676 679 4511 Total KPK 8010 13537 17805 24483 26613 30699 31108 34521 34275 221051

Figure 8: Annual Case Notification from 2002 to 2010 40000 34521 34275 35000 30699 31108 30000 26613 24483 25000 20000 17805 15000 13537 10000 8010 5000 0 2002 2003 2004 2005 2006 2007 2008 2009 2010

111

Figure 9: Total Case Notification by Districts in KPK, from 2002 to 2010

6.3 Annual Incidence of TB in KPK

The incidence rate of TB remained unchanged from 2002 to 2010, with an incidence of 177/100,000. The DOTS programme initiated in KPK in 2002 with 85% population coverage increased to 100% in 2005. Between 2002 and 2005, the number of registered TB patients increased by about 3.00% annually. The total annual TB patient registration decreased by 1.54% on average between 2005 and 2009. The reason behind this decrease may be due to 100 % population coverage by the DOTS programme in

KPK (See Table 18 on page 111). There was a reported decline of 1% in the success rate (number of patients successfully completed treatment with or without bacteriological success) from 2002 to 2003 which then gradually increased annually to

112

95% in 2010. The data also shows a fluctuation in the cure rate (number of patients with bacteriological evidence of success after treatment completion); initially the cure rate decreased by 2% between 2002 and 2003. However, between 2003 and 2004, the cure rate rose to 82%. This was followed by a decline of 2% between 2004 and 2006.

Between 2006 and 2010 the cure rate increased gradually and reached a maximum of

89% (See Similarly, the number of deaths increased from 17 to 385 from 2002 to 2006 and then deceased gradually to 219 in 2010 (See Figure 11 on page 114).

Table 19: Incidence of TB from 2002 to 2010 in KPK, Pakistan

Pop Under DOTS % DOTS Under Pop

Success Rate % Rate Success

Default Rate % Rate Default

Failure Rate Rate Failure

Incidence Rate Incidence

Death Rate % Rate Death

Cure Rate % Rate Cure

Total Cases Total

Pulmonary Pulmonary

Population Population

In millions In millions

Total Ext. Ext. Total

Total Total Year

2002 20.46 8,010 85 177 5,892 2,118 86 79 9 1 1 2003 20.56 13,537 85 177 9,727 3,810 85 77 9 1 2 2004 20.56 17,805 85 177 13,229 4,576 89 82 6 1 2 2005 20.75 24,483 100 177 17,329 7,154 90 81 4 1 3 2006 21.16 26,613 100 177 19,203 7,410 92 80 3 1 2 2007 21.59 30,699 100 177 21,919 8,780 93 84 2 1 2 2008 22.02 31,108 100 177 22,601 8,507 93 85 2 1 2 2009 22.46 34,521 100 177 24,669 9,852 94 88 1 1 2 2010 25.23 34,275 100 177 23,762 10,513 95 89 1 1 2

Figure 10: Cure and Success Rates of TB in KPK, 2002-2010 95

75 2002 2003 2004 2005 2006 2007 2008 2009 2010

Cure Rate Success Rate

113

Figure 11: Total Number of Annual Deaths from TB in KPK, 2002-2010 450 400 350 300 250 200 150 100 50 0 2002 2003 2004 2005 2006 2007 2008 2009 2010

Number of Deaths

Three major trends can be seen for the default rate from the data reported between 2003 and 2010. Firstly, between 2002 and 2003 the default rate remained stable at 9%. Secondly, a gradual decline to 2% reported between 2003 and 2007.

Finally, the default rate remained stable from 2% to 1% between 2007 and 2010 (See

Figure 12 below). The increase in cure rate and decrease in the number of deaths between 2006 and 2010 may be due to the DOTS programme achieving full population coverage in 2005.

Figure 12: Annual Default Rate of TB Patients in KPK, 2002-2010 10 9 8 7 6 5 4 3 2 1 0 2002 2003 2004 2005 2006 2007 2008 2009 2010

Default Rate

114

6.4 Distribution of TB Patients by Age and Gender

From 2002 to 2010, 221,051 TB patients registered with the NTP in the province of KPK, Pakistan. Table 20 on page 116 describes the male and female distribution of TB patients in the specified period, contributing 43.78% and 56.22 % respectively to the registered population. A higher number of female patients registered throughout the reported period when compared to males. The differences in gender distribution were statistically significant. (Chi-Square=29.74, p= <0.0001). Between

2002 and 2009, there was a gradual increase in the number of TB cases reported both in males (3428 to 15269) and females (4582 to 19252). Between 2009 and 2010, the number of male TB patients increased from 15269 to 15473, while the number of female patients decreased from 19269 to 18802 (See Figure 13 on page 116).

6.5 Classification of TB in Patients

Tuberculosis can be classified into pulmonary TB (PTb) and extra pulmonary

TB (EXTb). NTP data from 2002 to 2010 showed that 71.60% (158331) were PTb and

28.40% (62720) were EXTb. EXTb patients increased from 2118 to 10513 between

2000 and 2010 while, PTb cases increased from 5892 to 24669 between 2002 and 2009 with a decline noted from 24669 to 23762 between 2009 and 2010 (See Figure 14 on page 116).

115

Table 20: Distribution of TB Patients by Gender and Type of Disease, 2002-2010

Pulmonary TB Extra Pulmonary TB TOTAL Year Male (%) Female (%) Total Male (%) Female (%) Total Male (%) Female (%) Total

2002 2517 (42.7) 3375 (57.3) 5892 911 (43.0) 1207 (57.0) 2118 3428 (42.8) 4582 (57.2) 8010

2003 4203 (43.2) 5524 (56.8) 9727 1652 (43.4) 2158 (56.6) 3810 5855 (43.3) 7682 (56.7) 13537

2004 5628 (42.5) 7601 (57.5) 13229 2051 (44.8) 2525 (55.2) 4576 7679 (43.1) 10126 (56.9) 17805

2005 7246 (41.8) 10083 (58.2) 17329 3188 (44.6) 3966 (55.4) 7154 10434 (42.6) 14049 (57.4) 24483

2006 8335 (43.4) 10868 (56.6) 19203 3391 45.8) 4019 (54.2) 7410 11726 (44.1) 14887 (55.9) 26613

2007 9355 (42.7) 12220 (55.8) 21919 3922 44.7) 5202 (59.2) 8780 13277 (43.2) 17422 (56.8) 30699

2008 9813 (43.4) 12788 (56.6) 22601 3819 (44.9) 4688 (55.1) 8507 13632 (43.8) 17476 (56.2) 31108

2009 10715 (43.4) 13954 (56.6) 24669 4554 (46.2) 5298 (53.8) 9852 15269 (44.2) 19252 (55.8) 34521

2010 10669 (44.9) 13093 (55.1) 23762 4804 (45.7) 5709 (54.3) 10513 15473 (45.1) 18802 (54.9) 34275

Figure 13: Distribution of TB Patients by Gender, 2002-2010

Male Female

19252 17422 17476 18802 14887 14049 10126 15269 15473 7682 13277 13632 4582 11726 10434 5855 7679 3428 2002 2003 2004 2005 2006 2007 2008 2009 2010

Figure 14: Distribution of TB Patients by Disease Type, 2002-2010

116

6.6 Notification of New Sputum Positive Cases by Age Structure

Table 17 shows the registered number of sputum smear positive TB patients

between 2002 and 2009 (See Table 20 on page 116). Case notification reports suggested

approximately 50% annual increase in the number of TB patients from 2002 to

2006.The biggest number (74.511%) of patients registered annually were between 15 to

44 years of age, the most productive age of life. Males contributed 30.65% and females

43.86% of this group. Figure 15 shows that greatest gender difference was in the 14-24

age group. In the over 65 age group the incidence of TB was greater in males compared

to females (See Figure 15 on page 118).

Table 21: Notification of Smear Sputum Positive Cases by Age Structure in KPK, 2002- 2009

Age/Se 0-14 15-24 25-34 35-44 45-54 55-64 65 & > TOTAL x

Year M F M F M F M F M F M F M F M F Total

2002 52 164 401 554 289 388 172 304 129 157 116 93 80 60 1239 1720 2959

2003 88 204 548 774 415 543 247 355 230 279 219 208 160 129 1907 2492 4399

2004 93 310 759 1067 509 754 343 448 244 358 252 293 279 209 2479 3439 5918

2005 151 523 993 1531 724 1121 468 716 347 494 396 405 394 291 3473 5081 8554

2006 222 591 1223 1671 872 1249 585 787 482 593 415 492 481 389 4280 5772 10052

2007 221 635 1392 2110 1007 1374 651 898 559 724 545 567 653 530 5028 6838 11866

2008 225 634 1508 2185 1028 1555 616 895 601 782 581 631 750 526 5309 7208 12517

2009 242 572 1543 2416 1100 1593 686 994 599 767 646 668 725 571 5541 7581 13122

The average mean age for males was 36.63 and for females 32.71 between

2002 and 2009 (this was calculated by multiplication of mean age of every age set

according to annual cases, adding the figures of all age sets and dividing by total annual

cases registered in that year).

Figure 16 on page 118 shows the population pyramid of all smear positive TB

patients reported to NTP between 2002 and 2010

117

Figure 15: Distribution of New Smear Positive Cases by Age Structure 25000

20000 20675

15000 14521

10000 9165 7345 6527 6227 5000 4927

0 0-14 15-24 25-34 35-44 45-54 55-64 65 & Above

Male Female Total Cases as per age

Figure 16: Population Pyramid of New Sputum Positive Cases, 2002-2010

Table 22 on page 119 shows the gender distribution of TB cases by districts from 2002 to 2010. Of the total new sputum positive cases (n=69,387), 42.16 % were males and 57.84% females. The majority of districts reported a higher incidence of TB in female patients as compared to males. Only four districts were identified where males were at higher risk than females: Karak 2007 (m=79, f=75), Kohat 2003 (m=15, f=12),

Nawshera 2003 (m=216, f=210) and Shangla 2004 (m=16, f=12).

118

Table 22: Notification of New Cases by Gender and District, 2002-2010 2002 2003 2004 2005 2006 2007 2008 2009 G.Total No District M F M F M F M F M F M F M F M F M F Total

1 Abbottabad 0 0 82 79 187 221 232 306 397 432 394 533 362 521 379 572 2033 (43.28) 2664 (56.72) 4697

2 Bannu 0 0 76 81 36 47 69 102 140 167 254 339 307 385 313 411 1195 (43.82) 1532 (56.18) 2727

3 Battagram 0 0 0 0 3 7 40 96 30 71 44 109 55 150 69 121 241 (30.31) 554 (69.69) 795

4 Buner 14 19 47 78 47 76 83 133 83 190 103 214 108 237 99 205 584 (33.64) 1152 (66.36) 1736

5 Charsadda 3 5 44 47 89 94 103 155 131 167 217 286 228 279 275 343 1090 (44.20) 1376 (55.80) 2466

6 Chitral 6 16 81 154 85 167 94 229 79 161 69 190 90 179 90 191 594 (31.58) 1287 (68.42) 1881

7 DI Khan 127 175 96 121 131 220 213 403 245 293 341 471 234 331 250 365 1637 (4076) 2379 (59.24) 4016

8 Hangu 0 0 0 0 0 0 43 53 24 35 38 65 58 111 101 118 264 (40.87) 382 (59.24) 646

9 Haripur 0 0 121 157 183 246 223 326 255 366 210 295 242 305 261 315 1495 (42.56) 2010 (57.35) 3505

10 Karak 0 0 0 0 0 0 48 50 65 95 79 75 120 125 154 166 466 (47.70) 511 (52.30) 977

11 Kohat 0 0 15 12 65 67 128 131 132 183 214 181 154 201 205 252 913 (47.06) 1027 (52.94) 1940

12 Kohistan 0 0 0 0 1 2 23 88 39 131 34 117 44 136 49 193 190 (22.17) 667 (77.83) 857

13 Lakki Marwat 0 0 0 0 16 23 123 147 132 147 99 162 150 177 133 161 653 (44.42) 817 (55.58) 1470

14 Lower Dir 0 0 15 22 35 61 78 124 127 194 155 199 163 197 171 183 744 (43.16) 980 (56.84) 1724

15 Malakand 0 0 2 5 33 43 45 41 73 84 90 107 123 158 165 204 531 (45.27) 642 (54.73) 1173

16 Mansehsra 0 0 32 50 150 177 147 220 305 477 308 460 282 354 334 472 1558 (41.35) 2210 (58.65) 3768

17 Mardan 228 244 145 162 362 391 302 389 285 293 450 532 486 563 501 546 2759 (46.93) 3120 (53.07) 5879

18 Nowshera 168 191 216 210 163 262 237 294 249 262 240 288 319 377 422 449 2014 (46.93) 2333 (53.67) 4347

19 Peshawar 453 637 504 602 557 768 584 721 732 918 788 1008 786 1028 755 995 5159 (43.59) 6677 (56.41) 11836

20 Shangala 0 0 0 0 16 12 114 135 121 192 120 139 103 151 108 184 582 (41.72) 813 (58.28) 1395

21 Swabi 17 24 119 192 102 212 141 290 265 342 223 330 286 457 273 481 1426 (37.99) 2328 (62.01) 3754

22 Swat 223 409 293 492 186 287 291 454 251 359 395 466 423 486 248 386 2310 (40.89) 3339 (59.11) 5649

23 Tank 0 0 0 0 2 2 57 81 85 119 90 155 92 135 95 138 421 (40.06) 630 (59.94) 1051

24 Upper Dir 0 0 19 28 30 54 55 113 35 94 73 117 92 167 91 130 395 (35.97) 703 (64.03) 1098 40133 Total KPK 1239 1720 1907 2492 2479 3439 3473 5081 4280 5772 5028 6838 5307 7210 5541 7581 29254 (42.16) 69387 (57.84)

119

Kohistan district showed the greatest gender difference with females constituting 77.80% of the cases compared to 22% for males. Another five districts,

Battagram (m=30.30%, f=69.70%), Chitral (m=31.60%, f=68.40%), Buner (m=33.60%, f=66.40%), Upper Dir (m=36.00%, f=64.00%) and Swabi (m=38.00%, f=62.00%) also reported statistics which showed that women accounted for more than 60% of the total

TB cases. Karak (m=47.70%, f=52.30%), Kohat (m=47.10%, f=52.90%), Mardan

(m=46.90%, f=53.10%), Nawshera (m=46.30%, f=53.70%) and Malakand (m=45.30%, f=54.70%) are the districts of KPK where number of female registrations were in lower than to the other districts (See Figure 17 and Figure 18on page 120 and 121).

Figure 17: Distribution of TB Patients by Gender and District, 2002-2010

Male Female

100% 90% 80% 70% 60% 50% 40% 30% 20% 10%

Tank

Swat

Karak

Kohat

Swabi

Buner

Bannu

Hangu

Chitral

Haripur

DI Khan DI

Mardan

Kohistan

Shangala

Lower Dir Lower

Peshawar

Upper Dir Upper

Malakand

Nowshera

Battagram

Charsadda

Mansehsra Abbottabad Lakki Marwat Lakki

120

Figure 18: Distribution of Patients Notified by Higher and Lower Districts

6.7 Summary

Chapter 6 has described the descriptive epidemiology of TB patients in the province of Khyber Pakhtunkhwa, Pakistan from 2002 to 2010. Results in this chapter confirmed that notification (and hence incidence rate) of female patients was greater than male patients. 121

CHAPTER 7: RESULTS - RISK FACTORS

7.1 Introduction

This chapter describes the results of selected risk factors responsible for the higher notification of Tuberculosis in the province of Khyber Pakhtunkhwa (KPK). The results relate to field study carried out in KPK, and compare the risk factors for higher versus lower notification districts identified in Chapter 6.

A total of 914 new patients were diagnosed as confirmed TB cases during the period 1st July 2012 to 30th September 2012 in ten selected districts of KPK. Of the

914 study subjects, 129 were children under 15 years and were excluded from the study.

The results therefore relate to 785 patients who were interviewed as part of the study.

The data were inputted using the Epi Info version7© developed by the Centre for Disease Control for epidemiological field analysis. Both Epi Info© and SPSS 22© were used for analysing the data. The relevant findings are presented in tabulated form with frequencies, percentages, mean, median and standard deviation. Chi-squire tests and multiple logistic regressions were used to examine correlation between selected variables.

122

7.2 Distribution of Patients by Districts

785 participants were diagnosed as confirmed TB patients during the 3 months period between 1st July 2012 and 30th September 2012. Of those registered, 41.15 %

(323) were males and 58.85 % (462) females. Table 20 shows the distribution of cases by district and gender in the higher notification districts compared to the lower notification districts. Significant gender differences were noted in TB cases from higher notification districts (Chi-Square= 3.29). In Kohistan 75.38% (65) of notified cases were females. Similarly, in Battagram, Chitral, Buner and Upper Dir, about 65% patients were females [Table 23].

Table 23: Distribution of Patients by Districts Male Female Total (n=323) (n=462) (n=785) Districts Frequency % Frequency % Frequency % Kohistan 16 24.62 49 75.38 65 8.28 Battagram 27 36.00 48 64.00 75 9.55 Chitral 35 36.84 60 63.16 95 12.10

Higher Higher Buner 24 36.92 41 63.08 65 8.28 Upper Dir 21 35.00 39 65.00 60 7.64 Karak 39 49.37 40 50.63 79 10.06

Kohat 42 48.84 44 51.16 86 10.96 Mardan 49 48.51 52 51.49 101 12.87

Lower Nawshera 40 45.45 48 54.55 88 11.21 Malakand 30 42.25 41 57.75 71 9.04

123

Figure 19: Gender Differences in TB Notification in Higher and Lower District

70 Male Female 60 50 40 30 20 10 0

7.3 Age Distribution of Patients

The mean age of the participants was 33.86 years with a standard deviation of

16.19. The ages of the youngest and oldest patients were 15 and 85 years respectively.

More than three-quarters (76.94%) of notified cases were below 45 years of age [Table

24].

Table 24: Distribution of Patients by Age

Age Group Frequency Percent

15 - <30 394 50.19% 30 - <45 210 26.75% 45 - <60 86 10.96% 60 - <75 77 9.81% 75 - <90 18 2.29% TOTAL 785 100.00%

124

7.4 Educational Status of Notified Cases in Higher versus Lower Notification Districts

The level of education was determined by self reports of level of schooling of notified cases. Table 22 shows that 79.30% of females were illiterate in higher notification districts compared to 38.00% in lower notification districts. Gender differences show that 29.30% of males completed high school education in lower districts compared to only 7.20% females in higher notification districts. The percentage of women who completed high school education in low notification districts was 29.8% compared to 7.2% in high notification districts (Male versus female in higher districts;

Chi-Square=70.34, p= <0.001) [Table 25].

Table 25: Educational Status of Patients by Higher against Lower Notification Districts Higher Lower (n=360) (n=425) Total Education Level Male Female Male Female (%) (%) (%) (%) (%) 44 188 82 86 400 Illiterate (35.80) (79.30) (41.00) (38.20) (51.00) 5 3 0 12 20 Madrassa (4.10) (1.30) (0.00) (5.30) (2.50) 23 26 41 37 127 Primary (18.70) (11.00) (20.50) (16.40) (16.20) 36 17 63 67 183 High School (29.30) (7.20) (31.50) (29.80) (23.30) 11 2 8 19 40 Secondary (8.90) (0.80) (4.00) (8.40) (5.10) 1 0 4 3 8 Graduation (0.80) (0.00) (2.00) (1.30) (1.00) 3 1 2 1 7 Post Graduation (2.40) (0.40) (1.00) (0.40) (0.90) Male vs female in Higher; Chi-Square=70.34, p= <0.001 Male vs female in Lower; Chi-Square=0.19, p= <0.663

125

7.5 Distribution of Patients by Family Size in Higher versus Lower Notification Districts

Table 26 shows that more than half (57.07%) of notified cases were living in overcrowded houses with 7-12 family members with little difference between high and low notification districts. Surprisingly, 24.45% of notified cases were living in severely overcrowded houses. The mean family size was 10 members per family with a range of

1 to 40 (SD 4.77). There was no significant difference noted in family sizes between higher and lower notification by districts (Chi-Square=2.27, df= 3 p= 0.519) [Table 26].

Table 26: Distribution of Notified Cases by Family Size in Higher and Lower Notification Districts Higher Lower Total Family (n=360) (n=425) (n-785) Size Frequency Percent Frequency Percent Frequency Percent

01-06 68 18.89% 93 21.88% 161 20.51% 07-12 204 56.67% 244 57.41% 448 57.07% 13 -20 74 20.56% 76 17.88% 150 19.11% >21 14 3.89% 12 2.82% 26 3.31% Chi-Square=2.27, df= 3 p= 0.519

7.6 Distribution of Patients by Income

To explore the level of poverty, participants were asked to identify approximate monthly family income. On the basis of household income, patients were divided into three groups. (Severely poor= < 5,000 Pak Rupees, Poor or Low Income=

< 10,000 Pak Rupees, High Income or Economically Stable = > 20,000 Pak Rupees, 1

US$= 102 Pak Rupees). Table 26 shows that nearly two thirds (63.18%) of notified cases were living below the poverty level, earring less than $100 a month. Only,

11.21% of notified cases were noted to be economically stable. These data need to be interpreted with caution because 30.09% of females were not sure about their monthly household income [Table 27]. 126

Table 28 shows that 41.77% females in higher notification districts were living in severe poverty compared to 26.67% of the females in the lower notification districts.

Thirty per cent of female notified cases in lower notification districts were economically stable compared to 10.84% in higher notification districts. The monthly income of female participants in higher notification areas was significantly different from the monthly income of female participants in lower notification areas (Higher versus lower;

Chi Square = 27.54, p=<0.001, df = 4, p=<0.001) [Table 28].

Table 27: Income Distribution of Notified Cases by Gender Monthly Income Male Female Total Level (n=323) (n=462) (n=785) (Pak Rupees) Frequency Percent Frequency Percent Frequency Percent 1 US $ = 102 PKR < 5000 100 30.96% 159 34.42% 259 32.99% < 10000 126 39.01% 111 24.03% 237 30.19% < 20000 26 8.05% 38 8.23% 64 8.15% > 30000 9 2.79% 15 3.25% 24 3.06% Not Sure 62 19.20% 139 30.09% 201 25.61% Male versus Female; Chi Square = 23.77, p=<0.001Table 28: Income Distribution of Notified Cases in Higher and Lower Notification Districts Monthly Income Higher Lower Level Total (Pak Rupee) Male Female Male Female (%) Total Total 1US $ = 102 PKR (%) (%) (%) (%) 38 99 62 60 259 < or 5000 137 122 (30.89) (41.77) (31.00) (26.67) (32.99) 46 42 80 69 237 < or 10000 88 149 (37.40) (17.71) (40.00) (30.67) (30.19) 14 5 12 33 64 < or 20000 19 45 (11.38) (2.11) (6.00) (24.67) (8.15) 4 3 5 12 24 > or 30000 7 17 (3.25) (1.27) (2.50) (5.33) (3.06) 21 88 41 51 201 Not Sure 109 92 (17.07) (37.13) (20.50) (22.67) (25.61) TOTAL 123 237 360 200 225 425 785 Higher versus lower; Chi Square = 27.54, p=<0.001 Male versus female in higher; Chi Square = 40.94, p=<0.001 Male versus female in lower; Chi Square = 13.19, p=0.01

127

7.7 Distribution of Notified Cases by Age of the Youngest Child

One of the risk factors that we were keen to explore was the impact of having young children on the notification rate in women between high and low notification districts. This may give us an indication of nutritional status of notified cases. Our data showed that 38.60% male and 61.40% female notified cases were married. Of those who were married, 22.85% of notified cases in higher notification districts had a child who was less than a year old compared to 15.08% in lower notification districts., The differences between high and low notification districts were not significant.

Table 29: Distribution of Notified Cases by Age of the Youngest Child Higher Lower Total Youngest Child (n=302) (n=303) (n=605) Age Frequency Percent Frequency Percent Frequency Percent

1-3 Months 24 7.95% 22 7.26% 46 7.60% 6-12 Months 45 14.90% 24 7.92% 69 11.40% 1-2 Years 97 32.12% 62 20.46% 159 26.28% 3-4 Years 30 9.93% 64 21.12% 94 15.54% 5-10 Years 16 5.30% 42 13.86% 58 9.59% Above 10 63 20.86% 69 22.77% 132 21.82% Years No Child 27 8.94% 20 6.60% 47 7.77% Higher vs. lower; Chi-Square= 5.48, p=0.091

7.8 Distribution of Notified Cases by Type of Fuel Used for Cooking

The type of fuel used for cooking may be an indication of poverty and also of the environment. The later may also be a risk factor for TB as discussed in Chapter 3

Table 30 on page 129 shows that the vast majority (80.38%) of notified cases were using firewood as the main source of fuel for cooking. In higher notification districts,

95.00% (342) of the notified cases were using firewood compared to 68.00% (631) in lower notification districts. There were also marked differences in relation to the use of 128 gas for cooking. The differences between notified cases in high and low notification districts in type of fuel used was significant (Chi-Square=81.14, p=<0.001) [Table 30].

Table 30: Distribution of Notified Cases of Fuel Used for Cooking. Higher Lower Total Fuel (n-=360) (n=425) (n=785) Frequency Percent Frequency Percent Frequency Percent Firewood 342 95.00% 289 68.00% 631 80.38% Gas 18 5.00% 125 29.41% 143 18.22% Other 0 0.00% 9 2.12% 9 1.15% Higher vs. lower; Chi-Square=81.14, p=<0.001

7.9 Distribution of Patients by House Type

Table 31 shows that more than three quarters (76.11%) of notified cases high notification districts were living in mud houses, with 11.76% living in brick houses.

This compares to 46.86% of cases living in mud houses and 21.8% in low notification districts. These differences were significant (chi-square=30.26, OR=3.049, p=<0.001)

[Table 31].

Table 31: Distribution of Notified Cases by House Type Higher Lower Total House Type (n-=360) (n=425) (n=785) Frequency Percent Frequency Percent Frequency Percent Mud House 274 76.11% 199 46.82% 473 60.25% Bricks House 42 11.67% 93 21.88% 135 17.20% Stones House 2 0.56% 15 3.53% 17 2.16% Chi-square=30.26, p=<0.001

7.10 Distribution of Notified Cases by Amount of Time Spent Outside the House.

The amount of time spent outside the home may be a risk factor for TB because it could reflect on issues such as Vitamin D deficiency and also on the level of emancipation of women – the suggestion being that women who spent more time at home were not given the freedom to take part in activities outside the home. Forty eight

129 per cent of women indicated that they were going out for at least one hour in a 24 hours period, with 33.98% going out for 2 to 5 hours. In contrast, 60.3% of men indicated that they spent 6 to 8 hours outside the home with 20.12% indicated that they spent more than nine hours in 24 hours. Only 3.72% of men reported that they were staying out for at least one hour. These differences were significant (Chi-Square=133.1, df = 3, p=<0.001) [Table 32].

Table 32: Gender Distribution of Notified Cases by Amount of Time Spent Outside the Home Hours Male Female Total spent (n=323) (n=462) (n=785) outside the Frequency Percent Frequency Percent Frequency Percent home 0 to 1 12 3.72% 222 48.05% 234 29.81% 2 to 5 51 15.79% 157 33.98% 208 26.50% 6 to 8 195 60.37% 79 17.10% 274 34.90% > 9 65 20.12% 4 0.87% 69 8.79% Chi-Square=133.1, p=<0.001

Table 33 shows that there was a difference in time spent outside the home between men and women living in high compared to low notification districts. These differences were significant (Chi-Square= 44.92, p=<0.001).

Table 33: Distribution of Notified Cases by Amount of Time Spent Outside the Home between High and Low Notification Districts Stay Higher (n=360) Lower (n=425) Total Out Male Female Male Female (%) Time Total Total (%) (%) (%) (%) 6 142 80 234 0 to 1 148 6(3.00) 86 (4.88) (59.92) (35.56) (29.81) 15 77 80 208 2 to 5 92 36(18.00) 116 (12.20) (32.49) (35.56) (26.50) 83 18 61 274 6 to 8 101 112(56.00) 173 (67.48) (7.59) (27.11) (34.90) 19 0 4 69 > 9 19 46(23.00) 50 (15.45) (0.00) (1.78) (8.78) Chi-Square= 44.92, p=<0.001

130

7.11 Summary Table of Gender Differences and Risk Factors for TB (Model-1)

Table 34 summarises the risk factors for TB by gender. In addition to the risk factors discussed in the preceding paragraphs, the table also includes data on the occurrence of anaemia and weight. There were significant gender differences as exemplified by the odd ratio for all the listed variables. This was the first stage in developing a logistic regression model for assessing the different contribution of the variable to the notification of TB.

131

Table 34: Gender Difference and Risk Factors for TB (Model-1) Male Female 95% CI Variables OR p-vale n(%) n(%) Lower Upper Education Uneducated 126(31.8) 270(68.2) 2.088 1.562 2.792 <0.001 Educated 197(50.6) 192(49.4) Districts Higher 123(38.08) 200(61.92) 0.584 0.44 0.78 <0.001 Lower 237(51.30)) 225(48.70) Income Status 1Low Income 162(35.2) 298(64.8) 1.806 1.352 2.412 <0.001 2High Income 161(49.5) 164(50.5) HB Level 3Anaemic 14(31.9) 299(68.1) 2.398 1.792 3.208 <0.001 4Non-Anaemic 183(52.9) 163(47.1) BMI 5Underweight 65(25.1) 194(74.9) 2.873 2.068 3.992 <0.001 6Non-underweight 258(49.0) 268(51.0) Age Younger <45 220(36.4) 384(63.6) 2.305 1.644 3.231 <0.001 Older >45 103(56.9) 78(43.1) Marital Status Married 235(38.6) 374(61.4) 1.591 1.136 2.23 <0.007 Unmarried 88(50.0) 88(50.0) TB Type Pulmonary 260(41.5) 366(58.5) 1.082 0.7588 1.544 0.731 Extra Pulmonary 63(39.6) 96(60.4) Family Size Ideal Family <5 32(35.2) 59(64.8) 0.7252 0.4577 1.149 0.171 Overcrowded >5 291(41.9) 403(58.0) Family Type Joint Family 295(42.7) 396(57.3) 1.756 1.101 2.801 0.016 Separated Family 28(29.8) 66(70.2) Cooking Mood Indoor 243(41.5) 343(58.5) 1.054 0.7595 1.462 0.756 Outdoor 80(40.2) 119(59.8)

1 Low Income = < 10,000 Pak Rupees/month 2 High Income = > 20,000 Pak Rupees/month 3 Anaemic = HB < 12.00gm/dl (Women), HB <13 gm/dl (Men) 4 Non-Anaemic = HB > 12 gm/dl (Women), HB > 13gm/dl (Men) 5 Underweight = BMI < 17.5 6 Underweight = BMI < 17.5 132

7.12 Summary Table of Risk Factors for TB Notification by High and Low Notification Districts. ( Model-2)

Table 35: Differences in Risk Factors for TB Notified Cases between Low and High Notification Districts Higher Lower 95% CI Variables OR p-vale n(%) n(%) Lower Upper Education Uneducated 228(57.6) 168(42.4) 2.642 1.99 3.53 <0.001 Educated 132(33.9) 257(66.1) Income Status 7Low Income 246(53.5) 214(46.5) 2.128 1.59 2.85 <0.001 8High Income 114(35.1) 211(64.9) HB Level 9Anaemic 201(45.8) 238(54.2) 0.993 0.75 1.32 0.962 10Non-Anaemic 159(46.0) 187(54.0) BMI 11Underweight 135(51.1) 129(48.9) 1.377 1.02 1.85 0.035 12Non-underweight 225(43.2) 296(56.8) Age Younger <45 278(46.0) 326(54.0) 1.03 0.74 1.44 0.87 Older >45 82(45.3) 99(54.7) TB Type Pulmonary 268(45.0) 327(55.0) 0.873 0.63 1.21 0.417 Extra Pulmonary 92(48.4) 98(51.6) Family Size Ideal Family <5 37(41.1) 53(58.9) 0.804 0.51 1.25 0.340 Overcrowded >5 323(46.5) 372(53.5) Persons Living/room Ideal Family <5 55(35.9) 98(64.1) 0.602 0.42 0.87 0.006 Overcrowded >5 305(48.3) 327(51.7) Family Type 13Joint Family 308(44.6) 383(55.4) 0.648 0.42 1.00 0.052 14Separated Family 52(55.3) 42(44.7) Cooking Mood Indoor 24(41.8) 341(58.2) 0.525 0.38 0.73 <0.001 Outdoor 115(57.8) 84(42.2)

7 Low Income = < 10,000 Pak Rupees/month 8 High Income = > 20,000 Pak Rupees/month 9 Anaemic = HB < 12.00gm/dl (Women), HB <13 gm/dl (Men) 10 Non-Anaemic = HB > 12 gm/dl (Women), HB > 13gm/dl (Men) 11 Underweight = BMI < 17.5 12 Underweight = BMI < 17.5 13 Joint Family = A cultural system of living of all family members including married brothers and parents together in a single house. 14 Separated Family = A family living without their family members (Brothers and sisters) in separate house after marriage 133

Table 35 shows that the significant differences between low and high notification districts were related to education level, income, BMI, overcrowding, family type and whether cooking was carried out indoors or outdoors.

7.13 Summary Table of Multiple Logistic Regression

The variables listed in Table 34 and 32 that were statistically significant were combined in a logistic regression model in order to assess the contribution of the variables to the notification of TB in high notification districts. The reference category was low notification Multi-variable analysis was performed using stata software by logistic regression with groups of Variables i.e. gender, employment, income status, anaemic status, age, stays out of home time and meat eating.

Notification = Gender + Education + Employment + Income + Anaemia +

Age + Stays out time + Meat eating + Family size + (Gender X Education) + (Gender

X Employment) +…………

The results are presented in Table 35. For gender there is a highly significant effect with women having an odds ratio of 8.7, suggesting that they have nearly 9 times higher odds of being in a high notification area than men.

A strong association was noted for the interaction between education and gender (OR = 0.16), suggesting that more educated women were around 6 times less likely to be in a higher notification area (1/0.16) compared to a low notification area. It was also observed from the logistic regression model that income level (OR = 0.42), anaemia (OR = 0.45) and unemployment (Or = 0.23) were also associated with being in a high notification district in the province of KPK. Lower notification districts with female gender were used as reference category for the model.

134

Table 36 shows that female patients in lower notification districts were six times (1/0.16) less likely to be educated when compared to females in high notification districts (OR= 0.16, p=<0.001). Similarly, women in lower notification districts were four times (1/0.23) less likely to be employed as compared to women in higher notification districts (OR= 0.23, p=<0.05). Female participants in lower notification districts were almost 2.5 times less likely (1/0.42) to have high income as compared to female patients in higher notification districts (OR= 0.42, p=<0.05). Similarly, older female patients (> 45 years) were 2.8 times (1/0.35) less likely to be in lower notification districts than females in higher notification districts (OR= 0.35, p=<0.05)

[Table 36].

135

Table 36: Summary Table of Multiple Logistic Regressions Odds lower Upper Variables Estimate Std err Z p-value ratio CI CI 5.7928 0.60 56.25 0.12988 (Intercept) 1.7566 1.1598 1.5146

15***Female 2.1644 0.3919 5.5230 8.7097 4.04 18.78 0.00000 1.3796 0.81 2.36 0.24042 Educated 0.3218 0.2741 1.1739 1.2855 0.74 2.23 0.37115 Employed 0.2512 0.2809 0.8943 1.0901 0.66 1.79 0.73438 High Income 0.0863 0.2542 0.3393

16**Non-anaemic 0.7164 0.2537 2.8240 2.0470 1.25 3.37 0.00474 1.4964 0.85 2.63 0.16246 > 45 Years 0.4031 0.2886 1.3969

***Female: -1.8059 0.3646 - 4.9527 0.1643 0.08 0.34 0.00000 Educated 17 *Female: -1.4290 0.7145 - 1.9999 0.2395 0.06 0.97 0.04551 Employed *Female: -0.8611 0.3477 - 2.4765 0.4226 0.21 0.84 0.01327 High Income *Female: -0.7932 0.3405 - 2.3296 0.4524 0.23 0.88 0.01983 Non-Anaemic Female: -1.0276 0.4168 - 2.4652 0.3578 0.16 0.81 0.01369 > 45 Years

15 ***; p = <0.0001 16 **; p = 0.001 17 8; p = 0.01 136

7.14 Summary

In this chapter, the analysis has explored different risk factors responsible for gender differences in TB in higher and lower notification districts of KPK.

The gender difference between higher and lower notification districts was found to be statistically significant. Similarly, the majority of female patients in higher notification districts were found to be illiterate, unemployed, poor and living in low socioeconomic status.

Finally, the multivariable analysis determined a highly significant effect with women having an odds ratio of 8.7 showing that they have nearly 9 times higher odds of being in a high notification area than men. The other largest effect is for the interaction of gender and education with more educated women being around 6 times less likely to be in a higher notification area.

137

CHAPTER 8: RESULTS - PATIENTS’ KNOWLEDGE

8.1 Introduction

This chapter describes the results to estimate the level of knowledge in relation to TB of newly diagnosed TB patients comparing patients in higher against lower notification districts.

8.2 Gender Distribution of Patients by Higher and Lower Notification Districts

Of the total 323 male participants, 38.08% (123) were registered from higher and 61.92% (200) from lower notification districts. Similarly, of the total of 462 female participants, 51.30% (237) were registered from higher and 48.70% (225) from lower notification districts. The differences in gender distribution between high and low notification districts were significant (Chi-Square=13.38, p=<0.001) [Table 37].

Table 37: Gender Distribution of Patients in Higher and Lower Notification Districts Male Female Total Case Frequenc Frequenc Frequenc Status Percent Percent Percent y y y Higher 123 38.08% 237 51.30% 360 45.86% Lower 200 61.92% 225 48.70% 425 54.14% TOTAL 323 100.00% 462 100.00% 785 100.00% Chi-Square=13.38, p=<0.001

138

8.3 Distribution of Patients by Urban and Rural Residence

The data in Table 38 shows that the majority 86.62% (680) of the participants were living in rural areas. Twenty three percent of (53) female participants from lower notification districts replied that they were living in towns compared to 9.28% (22) in higher notification districts. These differences were significant (Chi-Square=13.03, p=<0.001).

Table 38: Distribution of Patients by Urban and Rural Residence in Higher and Lower Notification Districts Higher Lower Total Urban-Rural Male Female Male Female (%) (%) (%) (%) (%) Township/ 9 22 21 53 105 Town- Urban (7.32) (9.28) (10.50) (23.56) (13.38) 114 215 179 172 680 Village- Rural (92.68) (90.72) (89.50) (76.44) (86.62) TOTAL 123 327 200 225 785 Chi-Square=13.03, p=<0.001)

139

8.4 Knowledge of Patients about the Causes of TB

When asked questions about the cause of TB, almost half of the participants provided the correct answer. However a large number of male (42.11%, n = 136) and female participants (57.58%, n = 266) were unaware of the causes of TB. There were significant differences between male and female patients in the higher and lower notification districts (Chi-Square= 21.40, df= 3, p= <0.001). Similarly, the differences between men in the high and low notification districts regarding the cause of TB were not significant (Chi-Square= 0.92, df= 3, p= 0.819). However, when I compared the results of women between high and low notification districts, I noted a highly significant difference (Chi-Square= 96.23, df= 3, p= <0.001) [Table 39].

Table 39: Knowledge of Patients about the Causes of TB by Gender in High and Low Notification Districts Male Female Cause of TB Higher Lower Total Higher Lower Total (%) (%) (%) (%) Germs

Bacillus 67 101 35 130 168 165 Bacteria (54.47) (50.50) (21.21) (87.79)

Mycobacterium 7 9 16 12 Hard Work 16 28 (5.69) (4.50) (6.75) (5.33) 0 3 2 1 Hereditary 3 3 (0.00) (1.50) (0.84) (0.44) 49 87 184 82 Don't Know 136 264 (39.84) (43.50) (77.64) (36.45) TOTAL 123 200 323 237 225 462 Male versus female: Chi-Square= 21.40, p= <0.001

Male versus male: Chi-Square= 21.40, p= <0.001

Female versus female: Chi-Square= 96.23, p= <0.001

140

8.5 Knowledge of Patients about the Transmission of TB

Table 40 summarises the results of the responses regarding the transmission of

TB. The majority of male and female patients 61.69% (285) replied correctly in relation to knowledge about the transmission of TB. However, 35.28% (163) of female patients and 24.46% (79) male patients did not know that TB is a transmittable disease.

Significant differences were noted between men and women in the higher and lower notification districts regarding knowledge about the transmission of TB (Male versus

Female: Chi-Square=10.46, df= 2, p= 0.005, Male versus Male: Chi-Square=11.36, df=

2, p= 0.003, Female versus Female: Chi-Square= 68.41, df= 2, p= <0.001) [Table 40].

Table 40: Knowledge of Patients about the Transmission of TB by Gender in High and Low Notification Districts Male Female TB Total Total Transmission Higher Lower Higher Lower (%) (%) (%) (%) 98 134 98 187 Yes 232 285 (79.67) (67.00) (41.35) (83.11) 7 5 9 5 No 12 14 (5.69) (2.50) (3.80) (2.22) 18 61 130 33 Don't Know 79 163 (14.63) (30.50) (54.85) (14.67) TOTAL 123 200 323 237 225 462 Male versus female: Chi-Square= 10.46, p= 0.005

Male versus male: Chi-Square=11.36, p= 0.003

Female versus female: Chi-Square= 68.41, p= <0.001

141

8.6 Knowledge of Patients about the Spread of TB

Nearly two-thirds of the participants were unaware as to how TB was spread.

Almost an equal proportion of male (31.89%) and female (30.95%) patients correctly reported that infectious droplet was the main cause of TB spread. There were no significant differences reported in the knowledge about the spread of TB between men and women (Chi-Square= 2.59, df= 4, p= 0.629). Similarly, the differences between men in the high and low notification districts regarding the spread of TB were not significant (Chi-Square= 7.39, df= 4, p= 0.117). However, when I compared the results for women between high and low notification districts, I noted a highly significant difference (Chi-Square= 27.58, df= 4, p= <0.001) [Table 41].

Table 41: Knowledge of Patients about the Spread of TB by Gender in High and Low Notification Districts Male Female TB Spread Higher Lower Total Higher Lower Total (%) (%) (%) (%) Infectious 46 57 50 93 droplet 103 143 (37.40) (28.50) (21.10) (41.33) spread Sharing 18 26 41 28 44 69 Food (14.63) (13.00) (17.30) (12.44) 1 11 6 3 Smoking 12 9 (0.81) (5.50) (2.53) (1.33) 4 5 4 10 Hard Work 9 14 (3.25) (2.50) (1.69) (4.44) 54 101 136 91 Don't Know 155 227 (43.90) (50.50) (57.38) (40.44) TOTAL 123 200 323 237 225 462 Male versus Female: Chi-Square= 2.59, p= 0.629

Male versus Male: Chi-Square= 7.39, p= 0.117

Female versus Female: Chi-Square= 27.58, p= <0.001

142

8.7 Knowledge of Patients about the Signs and Symptoms of TB

Over 40% of males (45.82%) and females (41.34%) correctly reported that having a cough of more than three weeks duration was the most common symptom of

TB. However, 13.31% male and 15.58% female participants still believed that fever was the most common symptom of TB. There was a significant difference reported regarding the knowledge of the signs and symptom of TB within females patients between higher and lower districts (Chi-Square= 24.50, df= 3, p= <0.001) [Table 42].

Table 42: Knowledge of Patients about Signs and Symptoms of TB by Gender in High and Low Notification Districts

Male Female TB Sign Higher Lower Total Higher Lower Total (%) (%) (%) (%) Cough more 56 92 72 119 148 191 than 3 weeks (45.53) (46.00) (30.38) (52.88) 21 22 46 26 Fever 43 72 (17.07) (11.00) (19.41) (11.56) 7 8 Chest pain 13(6.50) 20 11(4.64) 19 (5.69) (3.56) 39 73 108 72 Don't Know 112 180 (31.71) (37.50) (45.57) (32.00) TOTAL 123 200 323 237 225 462 Chi-Square= 24.50; df 3; p= <0.001

143

8.8 Knowledge of Patients about Drug Resistance of TB

Only 10.82% (50) of female and 13.31% (43) of male patients correctly reported that multi drug resistant (MDR) TB is the main complication of incomplete and inappropriate TB treatment. There was a significant difference noted between the knowledge of female patients regarding MDR TB in higher and lower notification districts (Chi-Square= 27.48, df= 3, p= <0.001). However, the difference in knowledge regarding MDR TB between male participants in higher and lower notification districts was not significant (Chi-Square= 4.31, df= 3, p= 0.230) [Table 43].

Table 43: Knowledge of Patients about Drug Resistance of TB by Gender in High and Low Notification Districts

Male Female Complications of Total Total TB Higher Lower Higher Lower (%) (%) (%) (%) Drug Resistance 13 30 10 40 43 50 TB (10.57) (15.00) (4.22) (17.78) 17 40 27 32 Prolong Disease 57 59 (13.82) (20.00) (11.39) (14.22) 18 29 57 31 Death 47 88 (14.63) (14.50) (24.05) (13.78) 75 101 143 122 Don't Know 176 265 (60.98) (50.50) (60.34) (54.22) TOTAL 123 200 323 237 225 462 Female: Chi-Square= 27.48, p= <0.001

Male: Chi-Square= 4.3, p= 0.230

144

8.9 Overall Knowledge of TB Patients

To assess the level of knowledge of TB patients, the scores for the knowledge questions were summated. The overall score was obtained by adding up all the correct answers. A mean score was calculated with patients having a score below the average classified as having poor knowledge and those with a score above the average being classified as having satisfactory knowledge. It can be seen from the data in Table 44 that female patients in lower notification districts were five times more knowledgeable than female patients in higher districts (Chi-Square=62.87, p= <0.001) [Table 44].

Table 44: Overall Knowledge of TB Patients by Gender in High and Low Notification Districts

Knowledge Male Female Total Total of TB (%) (%) patients Higher Lower Higher Lower (%) (%) (%) (%) 69 126 195 195 106 301 Poor (56.10) (63.00) (63.37) (82.28) (47.11) (65.15) 54 74 128 42 119 161 Satisfactory (43.90) (37.00) (39.63) (17.72) (52.89) (34.85) TOTAL 123 200 323 237 225 462

Female: Chi-Square= 62.87, p= <0.001

145

8.9.1 Patients’ Knowledge about the Cause, Symptoms,

Transmission and Spread of TB

Five questions were asked to assess the knowledge of patients regarding the causes, symptoms and transmission of TB. Of the female patients (462), 78.79% (130) in the Lower notification districts and 21.21% (35) in the higher notification districts replied correctly that germs, bacteria or mycobacterium cause TB. A significant difference was noted between the knowledge of females in lower and higher notifications districts about the causes of TB (Chi-Square=92.99, OR= 7.90, 95% CI=

5.06-12.33, p= <0.001). Table 45 summarises the results of patient's knowledge about the causes and spread of TB and shows that female patients in lower notification districts were more informed than female patients living in notification districts [Table

45].

8.9.2 Patients’ Knowledge about the Treatment and Complications

of TB

Table 46 summarises the responses to questions about the knowledge of TB patients in relation to perceptions, treatment and complications of TB. The majority, of patients (79.83% (95) in higher notification districts were unaware that TB treatment is free. Similarly, 68.10% (111) female participants in higher notifications districts were ignorant about how to take TB drugs.

Table 46 shows that female patients in higher notifications districts were more ignorant than females in lower notification districts regarding the duration, treatment and follow up of TB [Table 46].

146

Table 45: Summary of Responses of the Participants Regarding Causes, Symptoms and Spread of TB by Gender in High and Low Notification Districts Male Female Questions OR Chi-Square 95% CL p-value Higher Lower Higher Lower

Knowledge about cause, symptoms, transmission and spread of TB What germ causes TB? (Germ, Bacteria, Mycobacterium) 0.85 0.48 0.54-134 0.488 Correct 67(39.88) 101(60.12) 35(21.21) 130(78.79) 7.90 92.99 5.06-12.33 <0.001 Incorrect 56(36.12) 99(63.87) 202(68.01) 95(31.99)

Do you know the most common symptom of TB? (Cough, Cough more than 3 weeks) 1.0192 0.0086 0.65-1.60 1.000 Correct 56(37.84) 92(62.16) 72(37.70) 119(62.30) 2.5727 24.1115 1.76-3.76 <0.001 Incorrect 67(38.29) 108(61.71) 165(60.89) 106(39.11)

Do you know TB is Curable diseases? (Yes)

Correct 107(39.19) 166(60.81) 166(45.86) 196(54.14) 0.73 0.93 0.38-1.39 0.428 2.89 19.82 1.79-4.67 <0.001 Incorrect 16(32.00) 34(68.00) 71(71.00) 29(29.00)

Do you think TB is transmittable disease? (Yes )

Correct 98(42.24) 134(57.76) 98(34.39) 187(65.61) 0.52 6.05 0.30-0.88 0.015 6.98 85.17 4.52-10.77 <0.001 Incorrect 25(27.47) 66(72.53) 139(78.53) 38(21.47) How TB is spread? (Infectious Droplet) Correct 46(44.66) 57(55.34) 50(34.97) 93(65.03) 0.67 2.78 0.41-1.07 0.110 2.63 22.11 1.75-3.97 <0001 Incorrect 77(35.00) 143(65.00) 187(58.62) (132(41.38)

147

Table 46: Summary of Responses of the Patient's Knowledge about Treatment and Complications of TB by Gender in High and Low Notification Districts Male Female Questions OR Chi-Square 95% CL p-value Higher Lower Higher Lower Knowledge about treatment and complications of TB patients? What is the minimum duration of TB Treatment? (6 Months) 1.0088 0.0015 0.64-1.58 0.530 Correct 60(37.97) 98(62.03) 86(39.45) 132(6055) 2.4921 23.197 1.71-3.62 <0.001 Incorrect 63(38.18) 102(61.82) 151(61.89) 93(38.11) How you will take TB medicines? (Once a Day ) Correct 87(36.40) 152(63.60) 126(42.14) 173(57.86) 1.3103 1.0985 0.79-2.17 0.299 2.9309 28.451 1.96-4.38 <0.001 Incorrect 36(42.86) 48(57.14) 111(68.10) 52(31.90) Do you know TB diagnosis and treatment is free? (Yes) Correct 99(38.08) 161(61.92) 142(41.40) 201(58.60) 1.0008 0.0000 0.57-1.76 0.554 5.6030 52.233 3.41-9.21 <0.001 Incorrect 24(38.10) 39(61.90) 95(79.83) 24(20.17) Do you know about follow up of TB? Correct 9(47.37) 10(52.63) 7(24.14) 22(75.86) 0.6667 0.7386 0.26-1.69 0.466 3.5609 9.1367 1.49-8.51 0.003 Incorrect (114(37.50) 190(62.50) 230(53.12) 203(46.88) What is the main complication of incomplete or inappropriate TB treatment? (Multi drug resistance TB ) 1.4932 1.2957 0.75-2.99 0.312 Correct 13(30.23) 30(69.77) 10(20.00) 40(80.00) 4.9081 21.985 2.39-10.08 <0.001 Incorrect 110(39.29) 170(60.71) 227(55.10) 185(44.90) 148

Figure 20: Correct Responses of Female Patients about the Causes, Symptoms and Spread of TB 250

200

150 Higher 100 Lower

0 Cause Symptom Curable Transmission Spread Correct Responses of Female Patients

Figure 20 shows the summary of correct responses of female patients in higher and lower notifications districts regarding the causes, symptoms and transmission of

TB.

Figure 21: Correct Responses of Female Patients about the Treatment, and Complications of TB 250

Higher Lower 200

150

100

0 Duration Taking Cost Fallow up Complications Medicine Correct Responses of Female Pateints

Figure 21 shows the summary of correct responses of female patients in higher and lower notifications districts about the treatment, cost, fallow up and complications of TB. It again graphically demonstrates that women in higher notification districts had lower knowledge than women in low notification districts.

149

8.10 Summary

Chapter 8 describes the results of questions regarding knowledge of TB patients in the province of Khyber Pakhtunkhwa. The data show that female patients in higher notification districts were less knowledgeable than females in the lower notification districts. Knowledge may be one of the factors in explaining gender differences between high and low notification districts.

150

CHAPTER 9: RESULTS - PHYSICIANS’ KNOWLEDGE

9.1 Introduction

This chapter describes the results related to the level of knowledge of general practitioners working in selected TB diagnostic and treatment centres of KPK. These are the doctors who are responsible for the diagnosis, treatment and follow up of TB patients and assessing their knowledge is an important component of understanding why some areas may have higher notification rates

9.2 Demographic Distribution of Physicians

A total of fifty physicians were interviewed. Twenty five were selected from each of the higher and lower notification districts. All the participants included in the study were male. The mean age of the participants was 44.22 years, ranging from 32 to

55 years. There was no significant difference in between the mean ages of doctors working in higher (41.28 years) and lower districts (43.83 years) When categorized by those below 45 years and those above 45 years, there was no significant difference.(Table 44) (Chi-Square=0.080, p= 0.785) [Table 47].

Table 47: Age Distribution of Physicians in Higher and Lower Notification Districts Higher (n=25) Lower (n=25) Total Age Frequency Percent Frequency Percent Frequency Percent Younger 12 48.00% 11 44.00% 23 46.00% (<45) Older(>45) 13 52.00% 14 56.00% 27 54.00% Chi-Square=0.080, p= 0.785

151

9.3 Qualification and Service of Physicians

All the physicians were medical graduates but only 16.00% (8) had a postgraduate qualification [Table 48]. The year of graduation of doctors included in the study was ranged from 1980 to 2006.

Table 48: Distribution of Physicians by Qualification Qualification Frequency Percent MBBS 41 82.00% MBBS and Post 7 14.00% Graduation MD 1 2.00% MD and Post 1 2.00% Graduation TOTAL 50 100.00% More than half (54.00%) of the participants had less than ten years of service experience. There was no statistically significance reported between higher and lower districts and service experience (Chi-Square=0.080, p= 0.785) [Table 49].

Table 49: Distribution of Physicians by Service and District of Notification Higher Lower Total Service Frequency Percent Frequency Percent Frequency Percent < 10 Years 14 56.00% 13 52.00% 27 54.00% > 10 Years 11 44.00% 12 48.00% 23 46.00% TOTAL 25 100.00% 25 100.00% 50 100.00% Chi-Square=0.080, p= 0.785

9.4 Knowledge about the Causes, Spread and Diagnosis of TB

Only 58.00% (29) of general practitioners replied correctly that a cough of more than three weeks duration was the most common symptom of TB. Forty two per cent (21) correctly identified that cough of more than 2 weeks or combination of weight loss, fever, cough and anorexia were major signs and symptoms of TB. Nearly two- thirds of the physicians responded that sputum smear microscopy was the recommend 152 test for the diagnosis of TB. The remaining 40.00% (20) advised that sputum smear, microscopy with combination of x-ray chest, ESR (erythrocytes sedimentation rate) and biopsy. These results suggest that all the participants had satisfactory knowledge of causes, symptoms, spread and diagnosis of TB [Table 50].

9.5 Knowledge about Treatment of TB Patients

When asked the question "what is the minimum duration of TB treatment?” over three-quarters (38) of the physicians replied correctly describing the recommended duration of treatment specified by the NTCP and WHO. When asked "how will you treat a pregnant woman with TB?", only 14.00% (7) of the physicians gave the correct answer with 30% replying that they would treat a pregnant woman with TB in the same way as a non-pregnant female patient. Thirty per cent of the physicians failed to answer this question, leaving a blank response, suggesting that knowledge in relation to this question was limited. Nearly two-thirds (62.00%) of the respondents incorrectly identified the correct answer when asked about the recommended dosage of anti-TB drugs. Forty four per cent of these of these incorrect responses described drug dosage levels below the recommend guidelines. So although knowledge of the causes and symptoms of TB was good, nearly two-thirds of the doctors were less knowledgeable about the treatment and management of TB [Table 50].

9.6 Knowledge about Follow-up and Complications of TB

More than half (54.00%) of the physicians were aware about the DOTS programme. However, 60.00% (30) of the physicians were unaware about the guidelines regarding the correct follow-up of TB patients. When we asked "which TB drugs are contraindicated in Hepatitis?", only 24 (48%) of the doctors replied correctly. Most alarmingly, 88% (44) of the participants were ignorant about how to treat multi-drug

153 resistant (MDR) TB. So knowledge in relation to the follow-up and complications of

TB was poor [Table 50].

Table 50: Summary of Questions Asked for Assessing Physicians’ Knowledge about the Diagnosis, Treatment and Complications of TB. Correct response Incorrect responses Questions Frequency Percent Frequency Percent

Knowledge about cause, spread and diagnosis of TB

What germs cause TB? (Mycobacterium ) 50 100.00 0 0.00 Do you know the most common symptom of 29 58.00 21 42.00 TB?(Cough more than 3 weeks) How you will diagnose TB?(Sputum for AFB) 50 100.00 0 0.00 Do you think TB is a communicable disease? 50 100.00 0 0.00 (Yes) Do you know how TB is spread? (Infectious 50 100.00 0 0.00 Droplet) Knowledge about treatment of TB patients What is the minimum duration of TB treatment? (Cat-I=6months, Cat-II =8 38 76.00 12 24.00 Months) How you will plan TB treatment? (Initial 20 40.00 30 60.00 Phase and Continuation phase) How you will treat a pregnant woman with TB? (Avoiding Streptomycin and careful 7 14.00 43 86.00 monitoring) How you will treat a 15 year child with TB? 21 42.00 29 58.00 What are the recommended dosages of anti TB 19 38.00 31 62.00 drugs? Knowledge about follow up and complications of TB What is DOTS? 27 54.00 23 46.00 How you will follow up the TB patient? 20 40.00 30 60.00 Which TB drugs are contraindicated in 24 48.00 26 52.00 Hepatitis? What is MDR? (multi-drug resistance) 32 64.00 18 36.00

How you will treat MDR TB? 6 12.00 44 88.00

9.7 Overall Knowledge of Physicians

A total of 15 questions were asked to assess the level of knowledge of physicians regarding the causes, symptoms, diagnosis, treatment and special cases of

TB. In order to obtain an overall score related to knowledge, all the correct observations were marked one and incorrect answers given a zero score. The overall knowledge score was achieved by calculating the addition of all the correct responses. A mean score for

154 knowledge was derived (9.58) and physicians scoring below this mean were classified as having poor knowledge. Those with a score above this mean value were classified as having satisfactory knowledge. Overall, 54.00% (27) of the physicians were classified as having poor knowledge [Table 51].

Table 51: Knowledge of Physicians in Relation to Causes, Management and Complications and Summary of Overall Knowledge. Knowledge about Frequency Percent Complications of TB (n=50) Knowledge regarding the causes, symptom and spread of TB. Unsatisfactory 0 0.00% Satisfactory 50 100.00% Knowledge about the management of TB Unsatisfactory 30 60.00% Satisfactory 20 40.00% Knowledge about Complications and follow up of TB Unsatisfactory 32 45.45% Satisfactory 18 54.55% Over all knowledge of Physicians Unsatisfactory 27 64.00% Satisfactory 23 36.00%

155

When classified by higher and lower notification districts, there was no significant difference in overall knowledge of physicians (Chi-Square=0.08, p= 0.388).

These data suggest that it was not knowledge amongst physicians that accounted for differences in notifications between districts. An interesting finding in this study was that physicians trained by the National TB Control programme (NTCP) were significantly more knowledgeable than those who were not trained by the NTCP (Chi-

Square=15.52, p= <0.001) [Table 52].

Table 52: Summary of Overall Knowledge of Physicians by Age, Years of Service, Experience, Training and Qualifications. Knowledge P-value Variables OR Chi-Square Poor Satisfactory

Age 16(69.57) 7(30.43) 0.0486 Younger (<45) 0.30 4.15

Older (>45) 11(40.74) 16(59.26)

Service 8(34.78) 15(65.22) 0.014 < 10 Years 4.45 6.33

> 10 Years 19(70.37) 8(29.63)

TB experience 23(53.49) 20(46.51) 0.8742 Yes 0.86 0.033

No 4(57.14) 3(42.86)

Training 10(32.26) 21(67.74) 0.00007 Yes 17.85 15.52 No 17(89.47) 2(10.53)

Qualification 27(64.29) 15(35.71) 0.003 MBBS, MD 12.64 7.31

Post Graduate 1(12.50) 7(87.50)

Districts 14(56.00) 11(44.00) 0.388 Higher 1.17 0.08

Lower 13(52.00) 12(48.00)

156

9.8 Summary

Chapter 9 has described the results of knowledge of physicians working in the selected TB diagnostic and treatment centres and comparing the knowledge between those working in higher and lower notification districts of Khyber Pakhtunkhwa.

Although there were no significant differences in the overall knowledge between those working in high and low notification districts, the overall level of knowledge was generally poor re-enforcing the importance of training as a means of maintaining knowledge.

157

CHAPTER 10: DISCUSSION

10.1 Introduction

The aim of this research was to explore the risk factors associated with higher notification of TB in the province of Khyber Pakhtunkhwa (KPK), Pakistan. I was trying to identify why more women were presenting with TB in certain districts in KPK with a view to understanding how we can improve the associated morbidity and mortality of this condition.

Tuberculosis (TB) remains one of the most important causes of morbidity and mortality for young women in KPK. The literature review in Chapter 3 identified the fact that that globally, TB notifications are higher in men (compared to women).

However we have also established in Chapter 5, that within KPK women were at much greater risk of developing TB when compared to men. This is an unusual pattern which has not been previously documented and could have significant implications for women's health in the region (49)(112-114) and on the management of TB control programmes. There is an issue that needs to be considered in relation to the whether the data are an artefact – for example is the higher notification rate in the higher prevalence districts due to a deficiency in male cases or is there a true excess of female cases? A plausible explanation could be that a reduced male rate results from the fact that males work away from home and present in another district. They could of course be better nourished and hence less susceptible.

I have argued that the structure of the health system in Pakistan encourages use of the health centres where TB is diagnosed – the fact that diagnosis and treatment are free presents fewer barriers for accessing treatment. There is a consistent approach to the diagnosis and treatment of TB in these health centres and the trends in increasing notifications over the 10 year period suggest that patients are seeking out this service, 158 especially since the trend of notification is also plateauing. My own experience of working in this area suggests that there is not a migratory effect which can account for the under-reporting of males. However, I do believe that the fact that women are more under nourished as exemplified by being more underweight and having a higher incidence of anaemia suggests that socio-economic factors are probably responsible for some of the higher notification of women.

Data collected by the National TB program follows guidelines developed by the WHO and whilst these have not been independently verified, it is my view that the data are robust and comprehensive. I was previously employed by the organisation and can vouch for the completeness of data collection. It is unlikely that the higher notification rate for females is from under-reporting of males. The data in KPK in terms of gender notifications are distinctly different from the rest of Pakistan.

Although the higher notification of TB in women in KPK has been previously described (6)(92)(115), I believe that this is the first time that anyone has tried to establish the risk factors associated with this higher notification rate. It is through this understanding of the risk factors that services will be able to develop interventions aimed at mitigating the impact of TB in this group of patients. In addition, by exploring the knowledge of both patients and doctors we can consider the value of public health campaigns and training strategies to improve public understanding of the disease amongst this group of patients and on the training of doctors. , The reality remains that at present there are no specific interventions aimed at women. It is my contention that if the current stakeholders of the MOH, NTCP, and the WHO, who sponsor the TB control programmes, are to reduce the spread of TB then they have to consider the specific epidemiology of the disease in this area which has a much greater impact on women.

159

Before the discussion and comparison of the results of the field work with other published studies, it is important to discuss the methodology, study limitations and problems associated with this field work.

10.2 Discussions of Methodology and Limitations

The present study was carried out from 1st July 2012 to 30th September 2012 in 10 districts of Khyber Pakhtunkhwa (KPK), Pakistan. Districts were grouped into higher and lower on the basis of male to female notifications of TB that were identified from the National TB Control Programme (NTCP) data between 2002 and 2010. My plan was to randomly select five TB diagnostic and treatment centres (TBDCs) from each district. Initially, my research team (I include the technicians conducting the interviews and collecting the data here as well as my supervisory team) proposed the random selection of five Basic Health Units (BHUs) from each district, but after meeting with National TB Control Programme Officials, I was informed that diagnostic and treatment services for TB were not available in all BHUs and only designated

BHUs were providing those services. These were now designated as TBDC’s.

The first key issue faced in relation to the selection of the districts was that there were only four (TBDCs) in Kohistan district. After meeting with the district TB

Control Officer in Kohistan, I agreed that referred patients from the Rural Health Centre

(RHC) would be registered as TBD-RHC and that these patients would be treated as though they were from the fifth centre in the district. For the registration of patients, a trained medical technician was provided in much the same way as existed in the other four health centres. Data collected from TBD-RHC was further verified by the area supervisor who worked with the TB register of NTCP. The data related to the patients from TBD-RHC were matched with the TB register of NTCP.

160

In order to explore the risk factors for higher notification of TB in adult females in the province of KPK, I initially proposed to conduct research in districts with higher female notification against districts with higher male notification. However, in a descriptive analysis of NTCP data from 2002 to 2010, it was noted that all districts in

KPK reported higher female notification of women compared to men [Table 22]. Based on local knowledge, the research team suggested that we selected five districts

(Kohistan, Battagram, Chitral, Buner and Upper Dir) with the highest female notification and compared these with five districts (Karak, Kohat, Mardan, Nawshera and Malakand) with the lowest female notification when compared to males [Figure

17]. This could introduce a bias because the women with higher notification may have characteristics predisposing them to TB. I therefore re-read the data relating to this; however the baseline characteristics of both men and women in these districts suggest that there were no particular features that may make them more predisposed to TB, nor were there any features in the Health centres which might have resulted in increased notification of cases. In reality the control group became women in low notification districts and the logistic regression compared the features between the low and high notification districts in essentially the same geographic area of Pakistan. Some studies in the literature review described the importance of using appropriate control group as means of exploring risk factors for TB. In our view the comparison of patients in low and high notification areas was an appropriate comparator group. The use of newly diagnosed patients allowed us to identify a distinct group of patients who we could compare.

One of the issues with data collection was consent taking. Due to high level of illiteracy, written consent was a problematic issue and as agreed with the ethics committee, verbal consent was taken from all patients included in the Study. In many respects prior to carrying out the fieldwork I deliberated long and hard about this

161 process. I knew for instance that tape recordings would help with accuracy and transcribing (116). However I also know the region and the culture. Indeed, Hafkin &

Huryer found rural illiterate women in the region were often suspicious and reluctant to use technology (computers, tape recorders etc) (117). Also, Zakar et al cautioned on its use due to cultural belief. Video recording was not a viable option not only because of the aforementioned, but also because in some of the locations electricity was unpredictable (118). Therefore, medical technicians and LHWs were trained and informed to read the provided consent statement to all patients before starting interview for verbal consent. The training included communication skills and all aspects of sensitivity.

The exclusion criteria that we developed required us to exclude patients below the age of 15. In general, the date of birth and exact age awareness was very limited amongst rural populations where there is a high rate of illiteracy Therefore in order to ensure that we excluded people under 15 years of age we attempted to verify their age by cross referencing patients’ names on the TB register with NTCP. Similarly, many ages may be inaccurate with many patients reporting their approximate age. However I do not believe that this would have biased the overall results since the practice of using approximate ages is widespread and does broadly compare to peoples’ chronological ages.

Some researchers argue that the use of focus group discussion is a gold standard for assessing knowledge (119). However, due to cultural restrictions, geographical locations, and problems in transportation, time and accessibility to all patients, I used a cross-sectional survey for assessing the individual knowledge of participants.

In relation to knowledge assessment we know that patients with a previous history of TB or with a positive family history for TB are likely to be more 162 knowledgeable when compared to those without a family history or previous exposure.

These patients were not excluded when assessing such knowledge and this might have artificially raised the knowledge level of participants in the districts. However, the proportion of patients with such conditions was 9.01% (n=71), which we believe is not large enough to alter the knowledge level of the sample.

Because of social conventions regarding the role of women in KPK, female patients are not normally allowed to be interviewed by male interviewers or without being accompanied a family member. In this study, the National TB Control

Programme had to collect data and registration details of all newly diagnosed TB patients as part of the requirements of the National TB Programme. In such situations, women are unlikely to agree to give accurate information to the mainly male technicians working in these centres. To avoid this problem, I recruited lady health workers for both higher and lower notification districts to collect data from female patients, who refused to be interviewed by male medical technicians. Of the total 785 women, no one was reported as having declined an interview with a male medical technician.

Another limitation of my study design was my use of medical technicians already working in the TDBCs for data collection. Their presence possibly has led to professional bias in interviewing patients because they know the results will be published internationally and any refusal for interview or having very low knowledge will be reported to NTCP. However, in the training sessions in each district, all medical technicians were encouraged to be honest and accurate with the patient responses for data collection.

163

10.3 Discussions of Descriptive Epidemiology (2002-2010)

The objective of this part of the study was to confirm that female patients in

Khyber Pakhtunkhwa were at a higher risk of TB when compared to males. Using

NTCP data from 2002 to 2010, a preliminary quantitative analysis of the existing

National MoH data of confirmed TB admission records was performed to identify trends in TB notifications according to age, gender and location, Descriptive analysis of this data was critical in identifying the epidemiological characteristics of patients and planning for the data collection phase relevant for objectives two, three and four.

The data showed that a total of 222,051 TB patients were reported in KPK and a large number of cases were reported from Peshawar (5763, 16.81%) followed by

Mardan (3060, 8.93%) and Nawshera (2117, 6.18%). The lowest number of TB cases were reported in Tank (471, 1.37%) fallowed by Upper Dir (679, 1.98%). The difference in the registration of patients was more likely due to the population difference of people living within the districts [Table 7].

Of the total notifications in this data set, 43.78% were male and 56.22 % were female patients. A higher number of female patients were registered when compared to males throughout the reported period. None of the twenty-five districts had higher notifications of men. The differences in gender distribution were statistically significant

(Chi-Square=29.74, p= <0.0001). These gender differences were noted in research conducted by the National TB Control Programme in Pakistan, whose findings were similar to our results (6). These gender differences were also described by Sabawoon and Sato in Afghanistan, (6)(19)(49)(61)(120). The geographical proximity of these regions suggests that this gender difference is specific to this part of the world and could be linked to a range of factors including poverty, educational levels, and other factors

164 that relate to the highly gendered nature of the society where women’s status is very subservient to men.

165

10.4 Discussions of Results Risk Factors

10.4.1 Sample of Patients

Figures in Table 23 showed differences in the registration of patients between the selected districts. The difference in the percentages of patient registration was possibly due to the difference in the population of the districts and reflects population density [Table 7]. The majority (101) of patients were reported from Mardan Districts and sample size represents the figures reported in descriptive epidemiology of TB in

KPK, which validates our results [Figure 9]. This higher patient registration was probably due to the higher population of Mardan as compared to other districts.

The significant gender differences that were identified in our study (58% females compared to 41% males) contradicts the general pattern of TB notifications where there is a higher registration of men compared to women (72)(95)(112)(121).

10.4.2 Age Distribution of Patients

Studies in the literature review suggested that a large number of TB patients were in the reproductive age group (19)(72)(96)(114)(122). Our findings confirmed that over three-quarters (76.94%) of the participants were in a younger age group [Table 24].

This is a concern because those between the ages of 15 to 45 are considered to be the most economically active age group in the population, and morbidity due to TB in this group is a substantial economic burden for both the state and individuals. In a local study in Swat districts of KPK, researchers also found that younger female patients were more likely to be registered with TB when compared to males (120).

166

10.4.3 Educational Status of Patients

Our results suggest that the level of education of female patients could be related to the higher notification. Nearly eighty per cent (79.30%) of females were illiterate in higher notification districts compared to 38.00% illiteracy in lower notification districts (Chi-Square=70.34, p= <0.001) [Table 25]. Gender differences were striking - out of the total sample of illiterate patients (n=400), 68.50 % were female compared to 31.50% male. Possible causes for significant illiteracy in higher notification districts are geographical location of the area, low socio-economic status and lack of educational institutions in the area.

The logistic regression confirmed an association between being female and education with less educated women being around 6 times more likely to be in a higher notification area. These are highly significant and plausible results.

Literacy rates in general in Pakistan are higher in males than females (98).

Several studies in our literature review reported that TB occurrence is associated with education. Shah et al reported that more than half (n=239, 59.1%) of patients with TB in a juvenile prison were uneducated (112). Similar findings were reported by Hussain et al, where just over half (57.0%) of their study subjects were uneducated with a further third (34.0%) only receiving primary education. Similar results were reported by Akhter et al in a study of 170 patients where the majority (n=277, 71.9%) of participants were uneducated (68). Several other studies also reported the higher rates of illiteracy in females (54)(95)(79).

Mushtaq et al reported that the majority (25.5%) of patients from rural areas were illiterate. This study also suggested that patients living in urban areas were more educated than those living in rural areas (75). This is because of the availability of schools, better transportation system and socio-economic status of the people living in

167 urban areas. Our study was primarily based in rural areas, which again may explain the high illiteracy rates that we found

168

10.4.4 Socio-economic Status of Patients

Our results suggested that more than half (50.76%) of the patients were living in overcrowded houses with 7-12 family members. Shockingly, 24.45% of the participants were reported living with 13 to 40 family members in a single house.

Our results were not different from other studies that we identified in our literature review, which have suggested that overcrowding was a major risk factor for

Tuberculosis (54)(71)(75). The majority of patients living in higher notification districts were poor and living in a joint family system, which leads to overcrowding and limited space for living. Similarly, the income status is directly related to poverty and TB.

Soomro and Qazi, reported that nearly three quarters (71.2%) of the patients with TB belonged to low income groups (54). Similar findings were found by other researchers, where the majority of TB sufferers were classified as having low income (75)(77).

Our results therefore replicated what others have found where nearly two thirds

(63.18%) of the participants were living below the poverty level, earning less than $100 a month compared to the national average of $255 monthly (123-125). The monthly income of female patients in higher notification districts was significantly different from the monthly income of female participants in a lower group (Male versus Female: Chi square = 23.77, p=<0.001; Higher Versus Lower: Chi square = 27.54, p=<0.001)

[Table 27, Table 28]. Female participants in lower notification districts were twice as likely to be educated when compared to female patients in higher notification districts.

Similarly, female participants in lower notification districts were more likely to be employed than women in higher notification districts. Education and employment were therefore the two main reasons for the significant differences in income. The logistic regression confirmed the interaction between gender and education and gender and employment.

169

We also found a significant difference between the participants living in different types of houses in higher and lower notification districts (chi-square=30.26, p=<0.001) [Table 31].This difference is probably related to income and rurality of the districts. Due to financial limitations and transportation, people living in rural areas can't afford bricks and blocks, preferring mud and stones for construction (122)

Several studies have identified smoking as a major risk factor for Tuberculosis

(54)(72)(95). However in our study, out of the 462 notified cases, only 3 female participants reported smoking. Therefore, we didn’t find any association of cigarette smoking with higher notification of TB in females of Khyber Pakhtunkhwa. However,

(80.38%) of the participants were using firewood as a main source of fuel for cooking.

In higher notification districts 95.00% (342) of the patients were using firewood compared to 68.00% (631) in lower notification districts (Chi-Square=81.14, p=<0.001) [Table 30]. Indoor smoke from firewood has higher toxic poisons which could have increased the risk of TB infection in females in higher notification districts

(122)(126).

10.4.5 Malnourishment and Social Empowerment of Women

Findings in this study showed that the majority (68.10%) of female patients were anaemic. Similarly, 74.90% of female participants were underweight. Poverty and male dominant culture, where males have a higher priority than females, is common practice in Pakhtun's families who are prevalent in this geographical area. For example males are served food earlier than females who would also tend to get smaller portions of food. This inevitably contributes to malnutrition, which is one of the risk factor for

Tuberculosis (51).

Similarly, a significant difference was reported between male and female patients and the duration of time they spent out of the home (Chi-Square=133.1,

170 p=<0.001). Nearly fifty per cent (48%) (234) of female participants replied that they were only going out of the home for less than 1 hour per day. These figures suggest that females were not exposed to the same amount of sunlight as male participants, suggesting women may have lower levels of Vitamin D. In a study conducted in

Pakistan on 129 patients (male= 42.00%, female= 58.00%), 89.00% female participants were highly susceptible for TB infection due to vitamin D deficiency (127).

Furthermore, these figures also throw some light on the status of women. Female patients in lower notification districts were spending four times more time outside the home when compared to female patients in higher notification districts. One interpretation of the data could be that women in lower notification districts were more empowered in their social life when compared to female patients in higher notification districts. Agboatwalla et al also reported that females living in the rural areas were not allowed to go out of home freely (79). There is clearly a significant coalescence of educational level, income, empowerment and nutritional status and their association with TB.

10.4.6 Analysis of Logistics Regressions

A logistic regression was used to identify associations between gender difference in higher notification of TB and statistically significant variables reported in the univariate analysis. All significant variables (gender, education, income, etc) were included in the logistic regression model.

171 districts were almost 2.5 times less likely (1/0.42) to have high income as compared to female patients in higher notification districts (OR= 0.42, p=<0.05). Similarly, older female patients (> 45 years) were 2.8 times (1/0.35) less likely to be in lower notification districts than females in higher notification districts (OR= 0.35, p=<0.05)

Therefore the results in our study confirm that socio-economic factors were most likely risk factors associated with the higher notification of TB in female patients of KPK,

Pakistan. Education was the most important risk factor directly related to knowledge, employment, income and women’s empowerment, which were responsible for the higher notification of TB in KPK Province. Several studies in our literature review descried that socio-economic characteristics of an individual were the main risk factors for TB occurrence (68)(75)(77)(79)(128) so in that respect our findings confirm the important role of socio-economic factors as a risk factor for TB. One of the conclusions that we can draw from this data is that the increased notification rate for women could be because they are more susceptible – through malnutrition, anaemia, lack of Vitamin

D and overcrowding. Illiteracy is almost certainly a proxy for poverty. All these factors will contribute to a higher notification rate but in the absence of robust population level data I can’t conclusively state that women are at greater risk for contracting TB.

172

10.5 Patients’ Knowledge of TB

I believe that we have carried out the first study that has evaluated the knowledge of TB patients regarding TB in the province of KPK. The results of our study suggested that female patients in lower notification districts were six times more knowledgeable as female patients in higher districts. It is likely that the socio-economic factors identified above played a significant role on the level of knowledge of adult TB patients.

A significant gender difference was reported between the knowledge of patients in lower and higher notification districts about the causes, symptoms and spread of TB (Chi-Square= 13.20, p= <0.001). This means that male patients were more informed than females. However, female patients living in lower notification districts were more knowledgeable than female patients in higher notification districts about the causes and transmission of TB. Mushtaq et al also explored the knowledge and behaviour patterns about TB in urban and rural communities of Punjab province and reported that urban patients were more knowledgeable as compared to rural patients

(75).

Almost half of the participants in this study replied correctly that germs, bacteria or mycobacterium causes TB. However, the knowledge of 42 % of male and

57% of female participants was still poor (Chi-Square= 21.40, p= <0.001) [Table 39].

Nearly two-thirds of the participant's knowledge about the spread of TB was poor and the gender difference was statistically significant (Chi-Square= 27.58, p= <0.001). Ali et al reported that more than four out of five participants (n=166, 82.0%) were aware that TB was a communicable disease and that just over half (n=110, 54.0%) said that it was spread through respiratory droplets (78).

173

However, some false perceptions were also noted about the signs and symptoms of TB. For example, 13.31% of male and 15.58% of female participants still believed that fever was the most common symptom of TB. In a cross-sectional study of

170 patients, Khan et al also reported fever as a main symptom of TB in 57.70% of patients (77). Worryingly, 27% of male and 29% of female patients believed that sharing of food causes spread of TB. This is an alarming situation particularly for females living in rural areas, where they are already facing rejection from the community due to these false perceptions. This can only exacerbate the gender differences that we identified.

Even though treatment by directly observed treatment short course (DOTS) is free in Pakistan 79.83% (95) patients in higher notification districts were unaware that

TB treatment is free (Chi-Square=53.23, 95% CI= 3.41-9.21, p= <0.001). Knowledge of

TB patients was extremely deficient regarding treatment and follow up. Possible reasons for this higher ignorance of TB treatment may be due to poor implementation of

Community Health Worker (CHW) programme in higher notification districts. Several studies reported that DOTS and the role of CHW is essential in proper TB management

(121)(129-131).

Varmund et al reported that the number of MDR cases increasing in Pakistan is due to lack of community participation (19). Similarly, Javaid et al found that incomplete and inappropriate TB treatment may lead to increase in MDR TB in

Pakistan (77). So the lack of knowledge that we identified may be a factor in the gender differences we noted.

Results showed in this study that knowledge of female patients in higher notification was poor as compared to female patients in lower notification districts (Chi-

Square=62.87, 95% CI= 3.41-7.96, p= <0.001). There are several possible reasons for this poor knowledge of female patients in higher notification districts. Firstly, the 174 majority of female patients in higher notification districts were illiterate, unemployed and poor. Mondal et al reported that socio-economic indicators play an important role in the knowledge of TB patients (113). Therefore, due to low socio-economic status, the female patients’ knowledge is limited in higher notification districts. Secondly, despite

100% DOTS coverage in KPK, the programme is still inefficient in higher notification districts due to lack of proper monitoring and training of CHWs. Finally, geographical and culture issues are the main barriers for women's access to education and knowledge in higher notification districts.

175

10.6 Physicians’ Knowledge of TB

Overall, 54.00% (n= 27) of the general practitioners were reported with poor knowledge regarding causes, management and complications of TB

There was no significant difference reported between higher and lower notification districts and TB knowledge (Chi-Square=0.08, 95% CI= 0.39-3.58, p=

0.388). Results in this study showed that doctors who received training in TB were significantly more knowledgeable (Chi-Square=15.52, 95% CI= 3.44-29.69, p=

<0.001) [Table 52]. Several studies in our literature review reported similar results.

(80)(84)(132).

Ahmed et al in their study of 170 patients found that 68.0% of respondents correctly identified that a cough for ≥ 3 weeks was the main symptom of TB, with

32.0% saying that cough, together with haemoptysis and weight loss was one of the main features of TB (80). In another study in Pakistan, Khan et al reported that nearly all participants (95.0%, n=114) knew that cough was a major symptom of TB, whilst

86.7 % (n= 104) also said that fever and weight loss (74.2 %, n= 89) were features of the disease (84). Similarly, 40.00% of the participants recommend sputum smear microscopy with combination of x-ray chest, ESR (erythrocytes sedimentation rate) and biopsy for the diagnosis of TB. Whereas, World Health Organization and National TB

Control Programme recommend sputum smear microscopy alone for the diagnosis of

TB, our study findings are similar to other studies conducted locally and regionally on the same topic (80-81)(86)

Almost, two-thirds of the doctor's knowledge was poor regarding the treatment and management of TB. Over three quarters of the participants reported the correct duration for TB treatment. However, only 14.00% (n= 7) of the participants were aware of how to treat pregnant women who were diagnosed with TB. The majority (62.00%)

176 of the general practitioners were ignorant about the recommended dosage of anti TB drugs and 56.00% prescribed above the recommended dosage. A similar percentage was unaware of knowledge regarding follow up and complications of TB. Similarly, two thirds of the participants were ignorant about the follow up of TB patients. Several studies in our literature review also reported findings similar to our results (80-81)(84).

Since 1993, when TB was declared as a global emergency, DOTS has been recommended for treatment of TB. Follow up of TB patients is the most important component of TB DOTS for monitoring and treatment outcome. Ahmed et al found that

50.0% of the private practitioners who reported treating TB patients were not following

DOTS strategy. Furthermore, few knew how to follow up TB patients (80). Shah et al also reported similar findings in a research project to assess the level of knowledge and practices of PPs in two major cities of Punjab. Out of 245 PPs, nobody advised treatment under DOTS (83). Surprisingly, 88.00% (44) of the doctors were unaware about the management of multi drug resistant (MDR) TB. According to WHO and other studies, the prevalence of MDR TB increased in Pakistan (133-134) and this could be a factor. Varmund et al and Javaid et al found that incomplete and inappropriate TB treatment may lead to increase in MDR TB in Pakistan (19)(77). Therefore, knowledge of doctors regarding MDR TB is very important and needs to be assessed further.

177

CHAPTER 11: IMPLICATIONS, CONCLUSIONS AND RECOMMENDATIONS

This study is the largest study carried out to date to investigate the risk factors associated with higher notification of TB in Khyber Pakhtunkhwa (KPK), Pakistan. To investigate the issue, we used national data sources, field work and surveys of the knowledge of patients and physicians.

The objective of the first phase of the research project was to confirm the higher notification of TB in female patients in the province of KPK. Retrospective data from National TB Control Programme (NTCP) was collected for all registered TB patients between 2002 and 2010. Through descriptive analysis, we confirmed that there was a higher notification of female patients in Khyber Pakhtunkhwa.

The descriptive analysis showed that these gender differences were reported from 2002 to 2010 annually. Unfortunately, these findings have not been published and this has resulted in no action being taken to reduce the gender disparities. Therefore, it is recommended that annual TB data of the provincial TB Control Programme should be published scientifically for research purposes to monitor trends in the notification of TB cases and the gender differences.

Objective two of the research has explored different risk factors associated with the higher notification of TB in adult female patients of Khyber Pakhtunkhwa.

Results found in objective two showed that the majority of female patients in higher notification districts were illiterate, unemployed, poor and living with low socio- economic status. Furthermore, through logistic regression, education was identified as one of the main risk factors for the higher notification of female patients in KPK. The results are important because optimal TB control will only be possible if we have a better understanding of the gender-associated risk factors for TB.

178

Literacy level of females is much lower in Kohistan, Battagram and Upper Dir.

The Government of Pakistan should take necessary steps for increasing girls’ enrolment in education in higher notification districts of KPK. It is very encouraging that on the basis of our preliminary results, which were discussed with the Minster of Health

Khyber Pakhtunkhwa in 2012, the government started a financial support project

(PKRs.200/month) for girl students attending schools in Upper Dir, Battagram, Hango, etc. Because of the success of this program, a special increase in the financial support

(PKRs.1500-2000/month) was announced for girl students in Kohistan district (135).

In objective three, we investigated the knowledge of TB patients in the province of Khyber Pakhtunkhwa. We identified that female patients in higher notification districts were less knowledgeable than females in the lower notification districts.

It is recommended that special focus should be given to DOTS and Community

Health Workers (CHW) programmes. Both should be strengthened and integrated to provide health education to TB patients in KPK. Moreover, the Government of

Pakistan, National TB Control Programme and International donors should construct a media policy for public awareness about TB.

Finally, knowledge of general practitioners working in the selected TB diagnostic and treatment centres of higher and lower notification districts of KPK was assessed in objective four. This study suggested that Physicians’ knowledge about the treatment, follow up and complications of TB was poor and there needs to be regular training as per WHO and NTCP guidelines for appropriate TB management.

The findings of this study reported that Physicians’ knowledge regarding multi drug resistant (MDR) TB was extremely poor. It is more likely that due to incomplete and inappropriate treatment, TB will continue to become more complex in the future in

179 the form of MDR in Pakistan and particularly in KPK province. Therefore, it is highly recommended that special attention should be given to the causes, diagnosis and treatment of MDR TB.

180

CHAPTER 12: FUTURE RESEARCH

Recent descriptive research conducted by NTCP in four provinces of Pakistan reported that female patients in KPK and Balochistan are at higher risk for TB than those in Punjab and Sindh provinces (6)(115). Therefore, on the basis of the results of this study, it is recommended that the same study should be carried out in higher (KPK and Balochistan) versus lower (Punjab and Sindh) notification provinces. This will help identify risk factors for higher notification of females; if the results replicate those that I found then I would suggest that the case for a Pakistan wide initiative for increasing the educational level of women as a means of reducing the incidence of TB would be compelling.

The knowledge of doctors is vital for the diagnosis and treatment of TB. There is only one study available on the knowledge, practice and treatment behaviour of general practitioners in KPK (81). However, that study was carried out on private practitioners and had very limited information with a low sample size (n= 88).

Similarly, the sample size (n= 50) of our study was also small and was further limited to male general practitioners only. It is therefore recommended to thoroughly evaluate the knowledge of those physicians (male and female) who are responsible for the management of TB patients in KPK working in either Basic Health Units or as private practitioners.

181

CHAPTER 13: REFERENCES

1. NTCP; National TB Control Programe Pakistan. Annual Report 2013: Islamabad;2013.[Accessed on 13 July 2014].Available from: http://ntp.gov.pk/uploads/NTP_Annual_Report_2013.pdf

2. NTCP; National TB Control Programe Pakistan. National Tuberculosis Program of Pakistan. Annual Report 2012: Islamabad;2012.[Accessed on 10 June 2013]. Available from: http://ntp.gov.pk/uploads/ntp_1392163200_ANNUAL_REPORT_2012.pdf

3. McHugh TD. World TB Day 2014: Reach the three million: a TB test, treatment and cure for all. Trans R Soc Trop Med Hyg. 2014;108(3):119–20.

4. NTCP; National TB Control Programe. Annual Report for Year 2009. Ministry of Health Islamabad;2009.[Accessed on 10 March 2011]. Available from: http://ntp.gov.pk/uploads/ntp_1369800770_ANNUAL_REPORT_2009.pdf

5. Abu-Amero kk. Status of anti tuberculosis drug resistance in Saudi Arabia. East Mediterr Heal Journal. 2002;8(4-5):664–70.

6. Dogar OF, Shah SK, Chugtai A a, Qadeer E. Gender disparity in tuberculosis cases in eastern and western provinces of Pakistan. BMC Infect Dis. 2012;12(1):244–51.

7. Grange JM, Zumla a. The global emergency of tuberculosis: what is the cause? J R Soc Promot Health. 2002;122(2):78–81.

8. Gupta D, Das K, Balamughesh T, Aggarwal AN, Jindal SK. ROLE OF SOCIO- ECONOMIC FACTORS IN TUBERCULOSIS PREVALENCE. Indian J Tuberc. 2004;51(1):27–31.

9. Davies PDO. Risk factors for tuberculosis. Monaldi Arch Chest Dis - Pulm Ser. 2005;63(1):37–46.

10. Liu JJ, Yao HY LE. Relationship between tuberculosis prevalence and socio- economic factors in China. Zhonghua Liu Xing Bing Xue Za Zhi. 2004;25(12):1032–4.

11. Ibrahim KM, Khan S, Laaser U. Tuberculosis control: current status, challenges and barriers ahead in 22 high endemic countries. J Ayub Med Coll Abbottabad. 2002;14(4):11–5.

182

12. World Health Organization. Global Tuberculosis Report 2104 [Accessed on 28 September 2015]. Available from: WHO/HTM/TB/2014.08

13. World Health Organization. Global Tuberculosis Report 2013 [Accessed on 10 January 2014]. Available from: http://www.who.int/tb/publications/global_report/en/index.html

14. WHO; World Health Organization. Global Tuberculosis Control 2010 [Accessed on 10 January 2014]. Available from: http://reliefweb.int/sites/reliefweb.int/files/resources/F530290AD0279399C1257 7D8003E9D65-Full_Report.pdf

15. World Health Organization). Global Tuberculosis Control 2009: Epidemiology, Strategy, Fianancing [Accessed on 02 April 2011]. Available from: WHO/HTM/TB/2009.411

16. World Health Organization. Global tuberculosis control 2008 - surveillance, planning, financing. WHO report. 2008. [Accessed on 02 April 2011]. Available from: WHO/HTM/TB/2008.393.

17. L Gillini and A Seita. Tuberculosis and HIV in the Eastern Mediterranean Region. East Mediterr Heal J. 2002;8(6):699–705.

18. World Health Organization. Global Tuberculosis Control: Surveillance, Planning, Financing -WHO report 2007 [Accessed on 02 April 2011]. Available from: http://www.who.int/tb/publications/2007/en/

19. Vermund SH, Altaf A, Samo RN, Khanani R, Baloch N, Qadeer E, et al. Editorial Tuberculosis in Pakistan : A decade of progress , a future of challenge. J Pak Med Assoc. 2009;59(4):1–8.

20. World Health Organization. Global tuberculosis report 2011 [Accessed on 11 June 2012]. Available from: WHO/HTM/TB/2011.16

21. NTCP; National TB Control Programe Pakistan. TB-Fee Pakistan. Current Status and the way forward:Islamabad;2008.[Accessed on 02 April 2011]. Available from: http://www.ntp.gov.pk/resource.php

22. Raviglione MC. The TB epidemic from 1992 to 2002. Tuberculosis. 2003;83:4– 14.

23. World Health Organization. Global Tuberculosis Control 2009: Epidemiology, Strategy, Financing. Tuberculosis. 2009. [Accessed on 02 April 2011]. Available from: WHO/HTM/TB 2010.7 183

24. indexmundi. Pakistan Age structure: 2015 [2015 Oct 30]. Available from: http://www.indexmundi.com/pakistan/age_structure.html

25. National Institute of Population Studies Islamabad. Pakistan Demographic and Health Survey 2012-13: Islamabad;2013 [Accessed on 02 April 2011]. Available from: https://dhsprogram.com/pubs/pdf/FR290/FR290.pdf

26. World Health Organization. Pakistan: WHO statistical profile. 2015 [Accessed on 17 October 2015]. Available from: http://www.who.int/gho/countries/pak.pdf

27. Infoplease. Country Profile Pakistan. 2014 [Accessed on 08 August 2015]. Available from: http://www.infoplease.com/country/pakistan.html

28. Khyber Pakhtunkhwa. Geography of Khyber Pakhtunkhwa: Peshawar;2015 [Accessed on 30 September 2015]. Available from: http://www.khyberpakhtunkhwa.gov.pk/aboutus/Geography.php

29. Governament of Pakistan. Population Socio-economic and Development Profile of Pakistan: Islamabad;2002. [Accessed on 02 July 2010]. Available from: http://www.pap.org.pk/population/pdf/population.pdf

30. Map of Khyber Pakhtukhwa: Peshawar;2015 [Accessed on 15 October 2015]. Available from: http://www.mqm.org/pak-maps/pakistan2010.htm

31. The Friday Times. Queen Elizabeth II in Swat (1961) [Accessed on 01 September 2015]. Available from: http://www.thefridaytimes.com/tft/queen- elizabeth-ii-in-swat-1961/

32. KPK. Governement of Khybar Pakhtunkhwa: Peshawar;2014 [Accessed on 08 August 2014]. Available from: http://www.khyberpakhtunkhwa.gov.pk

33. Marmot MG, Wilkinson RG. Social Determinants of Health. 1999;306. [Accessed on 08 August 2014] https://assets.publishing.service.gov.uk/media/57a08c29ed915d622c00117d/Kha n_Socialdeterminants.pdf

34. World Health Organization. Health System Profile: Pakistan. EMRO Regional Health Systems Observatory, Cairo; 2007 [Accessed on 20 May 2010]. Available from: http://apps.who.int/medicinedocs/documents/s17305e/s17305e.pdf

35. Malik MA, Gul W, Iqbal SP, Abrejo F. ORIGINAL ARTICLE COST OF PRIMARY HEALTH CARE IN PAKISTAN Muhammad Ashar Malik , Wahid

184

Gul *, Saleem Perwaiz Iqbal , Farina Abrejo. 2015;27(1):88–92.

36. Callen M, Gulzar S, Hasanain A, Rehman A, Khan Y, Mehmood MZ. Improving Public Health Delivery in Punjab , Pakistan : Issues and Opportunities. Lahore J Ecnomics. 2013;18(SE):249–69.

37. World Health Organization. Global Tuberclosis Report. 2013 [Accessed on 11 June 2014]. Available from: http://www.who.int/tb/publications/global_report/en/index.html

38. Pakistan Bureau of Statistics. Grotwth and Investment: Pakistan Economic Survay, 2012-13. Islamabad,Governement of Pakistan; 2013. [Accessed on 11 June 2014]. Available from: http://finance.gov.pk/survey/chapters_13/01- Growth%20and%20Investment.pdf

39. S. Ejaz Wasti. Highlights of the Pakistan Economic Survey 2012-13. [Accessed on 11 June 2014]. Available from: http://www.finance.gov.pk/survey/chapters_13/HGHLIGHTS%202013.pdf

40. World Health Organization. WHO REPORT 2003: Global Tuberculosis Control- Survillance, Planning, Financing.2003 [Accessed on 02 April 2011]. Available from: WHO/CDS/TB/2003.316

41. Owais A, Khowaja AR, Ali SA, Zaidi AKM. Pakistan’s expanded programme on immunization: An overview in the context of polio eradication and strategies for improving coverage. Vaccine. Elsevier Ltd; 2013;31(33):3313–9.

42. The Expanded Program on Immunization (EPI) P. CHILDHOOD IMMUNIZATION IN PAKISTAN.2015 [Accessed on 15 September 2015]. Available from: http://epi.gov.pk/index.php/139-2/

43. BBC. Pakistan to immunise people going abroad against polio. 2014 [Accessed on 28 October 2015]. Available from: http://www.bbc.co.uk/news/world-asia- 27292868

44. National TB Control Programe. Annual Report 2009: National TB Control Programme Ministry of Health Pakistan; 2009. [Accessed on 02 April 2011] Available from: http://ntp.gov.pk/uploads/ntp_1369800770_ANNUAL_REPORT_2009.pdf

45. NTCP; National TB Control Programe. UPSCALING TUBERCULOSIS CARE AND CONTROL: PROGRESS AND THE WAY FORWARD.Pakistan;2010. [Accessed on 02 April 2011] Available from: http://ntp.gov.pk/uploads/ntp_1369807855_ANNUAL_REPORT_2010.pdf

185

46. National TB Control Programe. Annual Report Summary 2008: Pakistan; 2008. [Accessed on 02 April 2011] Available from: http://ntp.gov.pk/uploads/ntp_1369800720_ANNUAL_REPORT_2008.pdf

47. World Health Organization. Global tuberculosis report 2013. [Accessed on 28 September 2014] Available from: http://www.who.int/tb/publications/global_report/en/index.html

48. Pakistan NTCP. Module for MBBS Students: Islamabad; 2014 [Accessed on 28 September 2011]. Available from: http://ntp.gov.pk/uploads/ntp_1369817904_CURRICULUM_FOR_MBBS_STU DENTS.pdf

49. Sabawoon W, Sato H. Sex Difference in Tuberculosis in Afghanistan: A National Cohort Study. Mycobac Dis. 2012;2(3):155–60.

50. Khadim, M.K.; Sarfaraz, J.; Masud TI. Factors affecting tuberculosis control : decision-making at the household level. JCollPhysicians SurgPak. 2003;13(12):697–700.

51. Ullah S, Shah SH, Kamal A, Begum N, Khan G. EXTRAPULMONARY TUBERCULOSIS IN LADY READING HOSPITAL PESHAWAR , NWFP , PAKISTAN : SURVEY OF BIOPSY RESULTS. J Ayub Med Coll Abbottabad. 2008;20(2):43–6.

52. Rizvi N, Shah RH, Inayat N, Hussain N. Differences in clinical presentation of pulmonary tuberculosis in association with age. J Pak Med Assoc [Internet]. 2003;53(8):321–4. Available from: http://www.scopus.com/inward/record.url?eid=2-s2.0- 0141682895&partnerID=tZOtx3y1

53. Israr SM. Is Ministry of Health fully prepared to implement an Effective DOTS Program in Pakistan? An Operations Research on TB Control Program in the Public Health Sector in Sindh. J Pak Med Assoc. 2003;53(8):324–7.

54. Shah S a, Mujeeb S a, Mirza a, Nabi KG, Siddiqui Q. Prevalence of pulmonary tuberculosis in Karachi juvenile jail , Pakistan. East Mediterr Heal J. 2003;9(4):667–74.

55. Rao NA, Sadiq MA. Recent trend in the radiological presentation of pulmonary tuberculosis in Pakistani adults. J Pak Med Assoc. 2002;52(11):501–3.

56. NTCP; National TB Control Programe Pakistan. Annual report 2006: Ministry of Health, Islamabd, Pakistan; 2006. [Accessed on 02 April 2011] Available from: http://ntp.gov.pk/uploads/ntp_1369800653_ANNUAL_REPORT_2006.pdf 186

57. NTCP; National TB Control Programe Pakistan. Anual Report 2007: Minsitery of Health, Government of Pakistan, Islamabad; 2007. [Accessed on 02 April 2011] Available from: http://ntp.gov.pk/uploads/ntp_1369800681_ANNUAL_REPORT_2007.pdf

58. World Health Organization. Global tuberculosis control 2011 - surveillance, planning, financing: WHO report; 2011. [Accessed on 02 April 2011] Available from: WHO/HTM/TB/2011.16

59. World Health Organization. Global tuberculosis report 2015. [Accessed on 12 October 2015] Available from: http://apps.who.int/iris/bitstream/10665/191102/1/9789241565059_eng.pdf

60. Nassaji M, Azarhoush R, Ghorbani R, Kavian F. Acid fast staining in formalin- fixed tissue specimen of patients with extrapulmonary tuberculosis. Int J Sci Res Publ. 2014;4(10):4–8.

61. Fader T, Parks J, Khan NU, Manning R, Stokes S, Nasir NA. Extrapulmonary tuberculosis in Kabul, Afghanistan: A hospital-based retrospective review. Int J Infect Dis. International Society for Infectious Diseases; 2010;14(2):e102–10.

62. Khan IM, Laaser U. Burden of Tuberculosis in Afghanistan : Update on a War- stricken Country. Croat Med J. 2002;43(2):245–7.

63. Roya Alavi-Naini, Batool Sharifi-Mood MM. The Gender Difference in Tuberculosis in Hihgly Endemic Region of Iran. J Med Sci. 2007;7(7):1218–20.

64. Chadha VK. Tuberculosis epidemiology in India: a review. Int J Tuberc lung Dis [Internet]. 2005;9(10):1072–82. Available from: http://www.ncbi.nlm.nih.gov/pubmed/16229217

65. Viney K, O’Connor J, Wiegandt A. The epidemiology of tuberculosis in Pacific Island countries and territories: 2000-2007. Asia-Pacific J public Heal. 2011 Jan;23(1):86–99.

66. Kakchapati S, Yotthanoo S, Choonpradup C. Modeling tuberculosis incidence in Nepal. Asian Biomed. 2010;4(2):355–60.

67. Kadivar MR, Ghaneh-Shirazi R, Khavandegaran F, Karimi M. Epidemiology of tuberculosis among Afghan immigrants in Far province, southren Islamic Republic of Iran. East MediterrHealth J. 2007;14(4):758–64.

187

68. Akhtar S, Rathi SK. Multilevel modeling of household contextual determinants of tuberculin skin test positivity among contacts of infectious tuberculosis patients, umerkot, Pakistan. Am J Trop Med Hyg. 2009;80(3):351–8.

69. Soomro JA QH. Factors Associated with Relapsed Tuberculosis in Males and Females: A Comparative Study. Tanaffos. 2009;18(3):22–7.

70. Aziz N, Hasan S, Munir M, Tayyab M, Chaudrhy N a. Risk to household contacts of tuberculous patients based on Mantoux test and antibody titre. J Ayub Med Coll Abbottabad. 2008;20(2):47–50.

71. Javaid A, Rizvi N, Ansari M, Sadiq A, Burki IS, Rehman NU, et al. Primary drug resistance to anti-tuberculous drugs in Karachi. J Coll Physicians Surg Pak. 2008;58(8):437–40.

72. Hussain H, Akhtar S, Nanan D. Prevalence of and risk factors associated with Mycobacterium tuberculosis infection in prisoners, North West Frontier Province, Pakistan. Int J Epidemiol. 2003;32(5):794–9.

73. Liefooghe R, Muynck a D. The dynamics of tuberculosis treatment adherence. J Pak Med Assoc. 2001;51(1):3–9.

74. Sharma PP, Kumar A, Singh P. A study of gender differentials in the prevalence of tuberculosis based on NFHS-2 and NFHS-3 data. Indian J community Med. 2010;35(2):230–7.

75. Mushtaq MU, Shahid U, Abdullah HM, Saeed A, Omer F, Shad MA, et al. Urban-rural inequities in knowledge, attitudes and practices regarding tuberculosis in two districts of Pakistan’s Punjab province. Int J Equity Health. BioMed Central Ltd; 2011;10(1):10–8.

76. Qureshi S a., Morkve O, Mustafa T. Patient and health system delays: Health- care seeking behaviour among pulmonary tuberculosis patients in Pakistan. J Pak Med Assoc. 2008;58(1):318–21.

77. Khan JA, Irfan M, Zaki A, Beg M, Hussain SF, Rizvi N. Knowledge, attitude and misconceptions regarding tuberculosis in Pakistani patients. J Pak Med Assoc. 2006;56(5):211–4.

78. Ali SS, Rabbani F, Siddiqui UN, Zaidi AH, Sophie A, Virani SJ, et al. Tuberculosis : do we know enough ? A study of patients and their families in an out-patient hospital setting in Karachi , Pakistan. Int J Tuberc Lung Dis. 2003;7(11):1052–8.

188

79. Agboatwalla M, Kazi GN, Shah SK, Tariq M. Gender perspectives on knowledge and practices regarding tuberculosis in urban and rural areas in Pakistan. East Mediterr Heal J. 2003;9(4):732–40.

80. Ahmed M, Fatmi Z, Ali S, Ahmed J, Ara N. Knowledge, attitude and practice of private practitioners regarding TB-DOTS in a rural district of Sindh, Pakistan. J Ayub Med Coll Abbottabad. 2009;21(1):28–31.

81. Shehzadi, R.; Irfan, M.; Zohra, T.; Khan, J.A.; Hussain SF. Knowledge regarding Management of Tuberculosis among General Practitioners in Northern Areas of Pakistan. J Pakistan Med Assoc. 2005;55(4):174–6.

82. Hussain A, Mirza Z, Qureshi F a., Hafeez A. Adherence of private practitioners with the national tuberculosis treatment guidelines in Pakistan: A survey report. J Pak Med Assoc. 2005;55(1):17–9.

83. Shah, S.K.; Sadiq, H.; Khalil, M.; Noor, A.; Rasheed, G.; Shah, S.M.; Ahmad N. Do private doctors follow national guidelines for managing pulmonary tuberculosis in Pakistan? East MediterrHealth J. 2003;9(4):776–88.

84. Khan J, Malik A, Hussain H, Ali NK, Akbani F, Hussain SJ, et al. Tuberculosis diagnosis and treatment practices of private physicians in Karachi , Pakistan. East MediterrHealth J. 2003;9(4):769–75.

85. Khan J a, Malik A. Tuberculosis in Pakistan: are we losing the battle? J Pak Med Assoc. 2003;53(8):320–3.

86. Rizvi N, Hussain M. Survey of Knowledge About Tuberculosis Amongst Family Physicians. J Pak Med Assoc. 2001;51(1):333–40.

87. Shen X, Deriemer K, Yuan Z, Shen M, Xia Z, Gui X, et al. Deaths among tuberculosis cases in Shanghai, China: who is at risk? BMC Infect Dis. 2009;9(3):95.

88. World Health Organization. Tuberculosis Control 2008: SURVEILLANCE, PLANNING, FINANCING. 2008. [Accessed on 02 April 2011] Available from: WHO/HTN/TB/2008.393

89. World Health Organization. Global Tuberculosis Report 2013. [Accessed on 28 september 2014] Available from: http://apps.who.int/iris/bitstream/10665/91355/1/9789241564656_eng.pdf

90. World Health Organization. Tuberculosis in women. 2014. [Accessed on 28 September 2014] Available from: 189

http://www.who.int/tb/publications/tb_women_factsheet_251013.pdf

91. World Health Organization. Gender and Tuberculosis (Gender and Health). 2002. [Accessed on 02 April 2011] Available from: http://apps.who.int/iris/bitstream/10665/68891/1/a85584.pdf

92. Codlin a. J, Khowaja S, Chen Z, Rahbar MH, Qadeer E, Ara I, et al. Gender Differences in Tuberculosis Notification in Pakistan. Am J Trop Med Hyg. 2011;85(3):514–7.

93. GBCHealth. WOMEN AND TB: HOW TO RESPOND. 2011. [Accessed on 13 June 2014] Available from: http://www.gbchealth.org/our- work/health_focus_areas/tuberculosis/

94. NTCP; National TB Control Programe Pakistan. Upscaling Tuberculosis Care and Control. Islamabad. [Accessed on 13 June 2014] Available from: http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=PMC4249520

95. Soomro JA, Qazi HA. Factors Associated with Relapsed Tuberculosis in Males and Females : A Comparative Study. Tanaffos. 2009;8(3):22–7.

96. Khattak MI, Ihsanullah, Muhammad A, Khan N, Zaman M. Frequency of sputum positive AFB cases among patients of pulmonary tuberculosis in tertiary care hospitals of northern Pakistan. J Ayub Med Coll Abbottabad. 2010;22(2):56–60.

97. Bassili A, Seita A, Baghdadi S, AlAbsi A, Abdilai I, Agboatwalla M, et al. Diagnostic and Treatment Delay in Tuberculosis in 7 Countries of the Eastern Mediterranean Region. Infect Dis Clin Pract. 2008;16(1):23–35.

98. Choudhry MA. Pakistan : where and who are the world ’ s illiterates ? [Internet]. 2006. Available from: 2006/ED/EFE/MRT/PI/22

99. Tuei VC, Maiyoh GK, Ha C, Sen T, Joshi SR, Udwadia ZF, et al. Tuberculosis: A Global Health Problem. BMC Infect Dis. 2013;13(7):122.

100. Aamir N. THE PLIGHT OF WORKING MOTHERS IN PAKISTAN : ADVANTAGES AND DISADVANTAGES OF A JOINT FAMILY SYSTEM [Internet]. 2004. Available from: http://www.spdc.org.pk/Data/Publication/PDF/CP-58.pdf

101. Gray A. Definitions of Crowding and the Effects of Crowding on Health : A Literature Review [Internet]. Wellington, New Zealand; 2001. Available from: http://www.msd.govt.nz/documents/about-msd-and-our-work/publications- resources/archive/2001-definitionsofcrowding.pdf 190

102. Wikipedia. Overcrowding. 2015 [Accessed on 30 October 2015]. Available from: https://en.wikipedia.org/wiki/Overcrowding#cite_note-who-1

103. Ignored LB. The Deadly Pairing of Tuberculosis and Smoking. World Lung Found. 2011;2(2):1–6.

104. Ramakrishnan C V, Rajendran K, Jacob PG, Fox W, Radhakrishna S. The role of diet in the treatment of pulmonary tuberculosis. An evaluation in a controlled chemotherapy study in home and sanatorium patients in South India. Bull World Health Organ. 1961;25(1):339–59.

105. Rahman AUR, Hayat Y, Habib Z, Iqbal J. Rural-Urban Disparities in Khyber Pakhtunkhwa Pakistan. Sarhad J Agric. 2011;27(3):477–83.

106. Ahmad SR, Velhal GD. Study of treatment outcome of new sputum smear positive TB cases under dots - strategy. Int J Pharma Bio Sci. 2013;4(3):B1215– 22.

107. Tuberculosis control programme of Khyber Pakhtunkhwa [Internet]. 2016 [cited 2016 Aug 4]. Available from: https://en.wikipedia.org/wiki/Tuberculosis_control_programme_of_Khyber_Pak htunkhwa

108. Provincial TB Control Programme [Internet]. 2016 [cited 2016 Aug 4]. Available from: http://ptp.gkp.pk/web/tbcenter

109. POLIT, DF & HUNGLER B. Nursing research: Principles and methods. 5th edition Philadelphia: Lippincott. 1995. p. 345.

110. Saunders M, Lewis P, Thornhill A. First Edition: Research Methods for Business Students. Edition. F, editor. Harlow, England: FT Prentice Hall.; 1997. 243 p.

111. Robson C. Real World Research. 3rd Edn. Chichester: John Wiley & Sons; 2011.

112. Shah SK, Dogar OF, Siddiqi K. Tuberculosis in women from Pashtun region : an ecological study in Pakistan. EpidemiolInfect. 2015;143(3):901–9.

113. Mondal MN, Nazrul HM, Chowdhury MR, Howard J. Socio-demographic factors affecting knowledge level of Tuberculosis patients in Rajshahi City, Bangladesh. Afr Heal Sci. 2014;14(4):855–65.

114. Bashour H, Mamaree F. Gender differences and tuberculosis in the Syrian Arab 191

Republic: patients’ attitudes, compliance and outcomes. East Mediterr Heal J. 2003;9(4):757–68.

115. Khan MS, Khan MS, Hasan R, Godfrey-faussett P. Unusual_sex_differences_in_tuberculosis notifications across Pakistan and the role of environmental factors. East Mediterr Heal J. 2013;19(9):821–5.

116. Kvale, S. & Brinkman S. Interviews: learning the Craft of Qualitative Research Interviewing. 2nd Editio. Beverly Hills, California: Sage Publications; 2009.

117. Neuberger J, Neuberger RJ. The patient / client / consumer / service user and medical ethics 40 years on. 2015;22–4.

118. Zakar R, Zakar MZ, Qureshi S, Fischer F. Harnessing information technology to improve women ’ s health information : evidence from Pakistan. 2014;1–13.

119. Eliot & Associates. Guidelines for Conducting a Focus Group [Internet]. 2005. Available from: https://assessment.trinity.duke.edu/documents/How_to_Conduct_a_Focus_Group .pdf

120. Ahmad Z. Prevalence of Tuberculosis in Tehsil Matta Swat Khyber Pakhtunkhwa. 2014;II(1):154–60.

121. Khan M. Tuberculosis patient adherence to direct observation: results of a social study in Pakistan. Health Policy Plan. 2005;20(6):354–65.

122. Singh AL, Jamal S. Unhealthy Cooking and Prevalence of Tuberculosis in Indian Women : A Case Study. J Environ Prot (Irvine, Calif). 2012;3(2):648–56.

123. Governament of Pakistan. HOUSEHOLD INTEGRATED ECONOMIC SURVEY ( HIES ) Statistics Division Pakistan Bureau of Statistics: Islamabad; 2013. [Accessed on 13 June 2014] Available from: www.pbs.gov.pk

124. Wikipedia. List of countries by average wage. 2015 [Accessed on 20 October 2015]. Available from: https://en.wikipedia.org/wiki/List_of_countries_by_average_wage

125. Governament of Pakistan. percentage distribution of monthly household income by source and quintiles, 2010-11. [Accessed on 13 June 2014] Available from: http://www.pbs.gov.pk/sites/default/files/pslm/publications/hies10_11/tables/tabl e11.pdf

192

126. Pokhrel AK, Bates MN, Verma SC, Joshi HS, Sreeramareddy CT, Smith KR. Tuberculosis and Indoor Biomass and Kerosene Use in Nepal: A Case–Control Study. Environ Health Perspect. 2009;118(4):558–64.

127. Talat N, Perry S, Parsonnet J, Dawood G, Hussain R. Vitamin D deficiency and tuberculosis progression. Emerg Infect Dis. 2010;16(5):853–5.

128. Hussain H, Akhtar S, Nanan D. Prevalence of and risk factors associated with Mycobacterium tuberculosis infection in prisoners , North West Frontier Province , Pakistan. Int J Epidemiol. 2003;32(5):794–9.

129. Dye C, Garnett GP, Sleeman K, Williams BG. Prospects for worldwide tuberculosis control under the WHO DOTS strategy. Lancet. 1998 Dec;352(12):1886–91.

130. Rajeswari R, Chandrasekaran V, Suhadev M, Sivasubramaniam S, Sudha G, Renu G. Factors associated with patient and health system delays in the diagnosis of tuberculosis in South India. Int J Tuberc Lung Dis. 2002;6(9):789–95.

131. Lu S-H, Tian B-C, Kang X-P, Zhang W, Meng X-P, Zhang J-B, et al. Public awareness of tuberculosis in China: a national survey of 69 253 subjects. Int J Tuberc Lung Dis. 2009;13(12):1493–9.

132. Udwadia ZF, Pinto LM, Uplekar MW. Tuberculosis management by private practitioners in Mumbai, India: Has anything changed in two decades? PLoS One. 2010;5(8):1–5.

133. World Health Organization. World Health Statistics: 2010 [Accessed on 20 May 2010]. Available from: http://www.who.int/gho/publications/world_health_statistics/EN_WHS10_Full.p df

134. Hasan R, Jabeen K, Ali A, Rafiq Y, Laiq R, Malik B, et al. Extensively Drug- Resistant Tuberculosis, Pakistan. Emerg Infect Dis. 2010;16(9):1473–5.

135. SHAH SQ. Stipend motivates girl students to continue education [Internet]. Dawn News. 2015. p. 3. Available from: http://www.dawn.com/news/1161230

193

CHAPTER 14: APPENDIX

14.1 Ethical Approval (The University of Manchester)

194

195

14.2 National TB Control Programme Ethical Approval Letter

196

197

198

199

14.3 Training of Medical Technicians for Data Collection Project

200

List of Medical Technicians Trained for data collection

Name Treatment Centre District Ghazi Khan TBC Daso Kohistan Fazal-u- Rehman TBC Daso -do- Khair Muhammad RHC Pattan -do- Umar Sadiq RHC Ranolai -do- Abdul Rashid RHC Shatyal -do-

Muhammad Amir BHU Shamlai Battagram Nasir Muhammad RHC Thakot -do- Muhammad Ayub THQ Bama -do- Hafeez Ullah RHC Koza Banda -do- Shakeel Ahmed TB Centre Battagram -do-

Ijaz Ahmed Khan DHQ/ Town/ RHC Kauzi Chitral Zakir Hussain Teh Torkoh/ Mulkoh -do- Muzfar Ahmed Chitral Rural -do- Subeiman ul Mulk Mastoch Sub Division -do- Altaf Ahmed Drosh -do- Amir Murad THQ Booni -do-

Abdul Haq DHQ Hospital Darrar Buner Bahram Shah THQ Hospital Pocho Kali -do- Mushtaq Ahmed Rural Health Centre Johar -do- Zafar Ali CH Totalai -do- Sher Rasool RHC Nagrai -do- Sher Zaman CH Chamla -do-

Safi ullah Khan DHQ Dir Upper Upper Dir Saiful Haq Cat D Hospital Patrak -do- Dr Wazir Zada RHC Bibyaware -do- Ghulam Hazarat Cat D Hospital Wari -do- Sadiqullah Cat D Hospital Barawal -do-

Noor Janan WCH Karak Karak Luqman uddin RHC Sabir Abad -do- Khalid Usman RHC Latamber -do- Type Chospital Takht e Muhammad Nasir -do- Nasrati Muhammad Iqbal DHQ Karak -do- Aziz ur Rehman RHC Ahmad Abad -do- List Continue....

Mir Nawab Khan DHQ Kohat Kohat Muhammad Yasin Liaqat MH Kohat -do- 201

Faisal Anwar RHC Lachi -do- Ghulam Haider RHC Chor laki -do- Muhammad Basheer RHC Ustarzai -do- Nazik Muhammad RHC Billitang -do-

Gul Sher Khan DHQ/DTO office Mardan Muhammad Ayub CH Lond Khwar -do- Muhammad Ali Type D Hospital Katlung -do- Aziz ur Rehman Mardan Medical Complex -do- Syed Sabir Shah CH Rustam -do- Muhammad Ali RHC Sheer Ghar -do-

Nasir Khan RHC Khair Abad Nawshehra Younas Khan CH Akora Khattak -do- Khiasta Rehman RHC Kheweshqi -do- Sharif Gul RHC Nizam Pur -do- Syed Sajjad Ali Shah CH Kaka Sahib -do- Majid Ali DHQ Naw Shehra -do-

Muhammad Ismail DHQ Batkhela Malakand Agency Muhammad Naeem Ch Thana -do- Mian Akbar ullah Ch Thota Kan -do- Ghour Ayub RHC Dhery Jolagram -do- Amar Nat DLS Malakand -do- Izat Begum RHC Skhakot -do-

202

Photographic presentation of Training Sessions of Medical Technicians

203

204

14.4 Data Collection Instruments

205

Islamic Development Bank National TB Control Programme, Saudi Arabia Khyber Pakhtunkhwa

S S

c c

h h

o o

l l

Tuberculosis Research Project o o

f f

Khyber Pakhtunkhwa (Pakistan) Questionnaires M M e e To Explore the Risk Factors for Tuberculosis in Adult TBd d i i

Patients of Khyber Pakhtunkhwa and to Assess the c c Knowledge of TB Patients about TB. i i n n e e , ,

S S t t o o

p p

f f

o o By Dr Muhammad Aziz r r

d d

B B

u u

i i

l l

d206 d

i i

n n

g g

Consent Statement

Informed consent for Tuberculosis research project in Khyber Pakhtunkhwa, Pakistan

Instruction to the interviewer: read this consent form in a loud and clear to the respondent

Good Morning/good afternoon. My name is [Name of the interviewer]. I am representing National Tuberculosis Control Programme and Ministry of Health Khyber Pakhtunkhwa who is carrying out some research in this Area, The University of Manchester, England, Islamic Development Bank and National Tuberculosis Control Programme are carrying out research on tuberculosis to improve the treatment and prevention of Tuberculosis in Khyber Pakhtunkhwa. Research indicates that female patients are more likely than males to have been diagnosed with TB in last 10 years in this province. The University of Manchester and the National Tuberculosis Control Programme would like to know what is happening in the general population in order to explore factors associated with this higher occurrence of tuberculosis in females. We will be doing this by asking you some questions and if you agree I will do this shortly. Some of these questions will be about your personal life. I am aware that some of the questions are sensitive, but all the information you give me will be kept strictly confidential. Participation in the survey is voluntary. You can refuse to answer all or some of the questions, but National Tuberculosis Control Programme and the University of Manchester would appreciate your help in answering all the questions. It is important to know that your participation will not affect your ability to use health facilities or free treatment. We are hoping that you will participate since your views are highly valued and important. Do you want to ask me anything about the survey? May I begin asking you question now?

[Instruction to the interviewer: Wait for the answer and make sure you do not rush the respondent into answering. If the respondent is not answering, gently ask the question again until you get an answer]

I understand the details of the survey have informed me about the requirements and hereby agree to participate in the survey.

The respondent has been informed about this survey and understands its purpose and objectives

Signature of the interviewer ______

* Modified consent form, form the Guidelines for measuring National HIV prevalence in population based surveys, 2005.

207

Tuberculosis Research Questionnaire

A. Baseline Data

1. Name ______2.Gender[1] Male [2]Female 2. 3. Age____ Years 4. Citizenship [1] Pakistani [2] Foreign Specify ______5.Weight__Kg 6.Height___Cm 7. BCG [1] No scar [2] Scar Seen [3] Dubious 8.HBLevel_____gm/dl 9. BHU / Treatment Centre______10.District ______11. Contact No ______12.Type of TB?[1] Pul [2] Ex Pul

13. Date of Entry ______14.Form No ______B. Interview I. Demographic Information

1. Educational Status [1] Illiterate [2] Primary [3] High School [4] Secondary [5] Graduate [6] Post Graduation [7] Madrassa [8] Other Specify ______

2. Occupation [1] None [2] House Wife [4] Teacher [3] Business [4] Driver [5] Labourer [6] Other Specify______

3. What is your average monthly income? [1] < 5000 [2] <10000 [3] <20000 [4] >30000 [5] Not sure [6] Refused to Answer

4. What is your marital status at time of Diagnosis of TB? [1] Never married [2] Married [3] Divorced [4] Widowed [5] Refused to Answer

5. For men only, how many wives do you have? [1] 1 [2] 2 [3] 3 [4] 4

6. How many Children do you have? [1] None [2] ______[3] Refused to Answer

7. Age of the youngest child? [1] 1-3 months [2] 6 months [3] 12 months [4] > 1 year

8. Family Type [1] Joint [2] Separated

9. Family Size [1]______Total family members sharing the same roof.

10. How many rooms do you have? [1] 1 [2] 2 [3] 3 [4] 4 [5] >5

11. How many people living per room? [1] 1 [2] 2 [3] 3 [4] 4 [5] >5

12. Cooking Mode [1] indoor [2] outdoor

13. Type of fuel [1] Firewood [2] Kerosene [3] Gas [4] Electricity

208

14. How many times you eat meat in a week? [1] 1 [2] 2 [3] 3 [4] other ______15. Place of residence [1] village [2] Township/ Town [3] Do not know

16. How many hours you stay out from home daily? [1] 0-1 [2] 2-5 [3] 6-8 [2] >9

17. Type of House [1] Mud House [2] Block / Bricks House [4] other Specify ______

18. Do you have toilet? [1] Yes [2] No

19. Placement of Toilet [1] Attached to Room [2] inside home [3] outside home

II. Past Medical History (Risk Factors)

20. Do you have any previous history of TB? [1] Yes [2] No [3] Refused to Answer

21. Does anyone in your family have history of TB? [1] Yes [2] No [3] Do not know

22. If yes who are affected? [1] Spouse [2] Parents (one or Both) [3] Brother/ Sisters [4] Children [5] Other Specify ______

23. Do you have any associated disease? [1] Blood Pressure [2] Diabetes [3] other Specify ______

24. Do you smoking? [1] Never [2] Past [3] Current

25. How many cigarettes per day? [1] 1-10 [2] 11-20 [2] >20

26. Do you snuffing? [1] Yes [2] No

III. Knowledge and Understanding of Tuberculosis

27. What Germ Cause TB? [1] Germ/ Bacillus [2] Hard Work [3] Hereditary TB [4] Do not know [5] Other ______

28. Do you know the most common symptom/sign of TB? [1] Fever [2] Cough more than 3 weeks [3] Chest pain [4] Do not Know [5] other Specify____

29. Do you think TB is curable disease? [1] Yes [2] No [3] Do not Know 209

30. Do you think TB is transmittable diseases? [1] Yes [2] No [3] Do not Know

31. How TB is spread? [1] Infectious droplet Spread [2] Sharing food [3] Smoking

32. What is the minimum duration of TB treatment? [1] 2 months [2] 6 Months [3] 8 Months [4] 2 years [5] Do not know

33. How you will take TB medicines? [1] once/day [2] Twice/day [3] Other

34. Do you know TB treatment is free? [1] Yes [2] No [3] Do not know

35. When is the follow-up performed? [1] At the end of 2nd, 4th and 6th Month [2] At the end of 2nd, 5th and 8th Month [3]After completion [4] Do not know

36. What is the main complication of incomplete or inappropriate TB treatment? [1] Prolong disease [2] Death [3] Drug Resistance TB [4] Do not know

210

Islamic Development Bank National TB Control Programme, Saudi Arabia Khyber Pakhtunkhwa

S S

c c

h h o o o o Tuberculosis Research Project l l

Khyber Pakhtunkhwa (Pakistan) o o f f

Questionnaire

M M

Knowledge of General Practitioners Regarding the Signs, e e Symptoms, Diagnosis and Treatment of TB Patients d d i i

c c i i

n n

e e

, ,

S S

t t

o o By Dr Muhammad Aziz p p

f f

o o r r

d d

B B

u u

i i

l l

211d d

i i

n n

g g Consent Statement

Informed consent for Tuberculosis research project in Khyber Pakhtunkhwa, Pakistan

Instruction to the interviewer: read this consent form in a loud and clear to the respondent

Good Morning/good afternoon. My name is [Name of the interviewer]. I am representing The University of Manchester United Kingdom. National Tuberculosis Control programme is working for the treatment and prevention of Tuberculosis in Khyber Pakhtunkhwa Research indicates that female patients are more likely to suffer from TB compared to males in last 10 years in this province.

The University of Manchester would like to know what is happening in the general population in order to explore factors associated with this higher occurrence of tuberculosis in females. We hope to gain this information by asking you some questions.

If you agree, I will ask you some questions. Some of these questions will be about your personal life. I am aware that some of the questions are sensitive, but all the information you give me will be kept strictly confidential. Participation in the survey is voluntary. You can refuse to answer all or some of the questions, but The University of Manchester would appreciate your help in answering all the questions. It is important to know that your participation will not affect your ability to use health facilities or free treatment. We are hoping that you will participate since your views are highly valued and important.

Do you have any questions about what I am asking you to do?

May I begin asking you question now?

[Instruction to the interviewer: Wait for the answer and make sure you do not rush the respondent to answer. If the respondent is not answering, gently ask the question again until you get an answer. If you feel that he/she is not willing to take part, don’t pursue this]

He/She understands the details of the survey have informed about the requirements and here by agree to participate in the survey.

ID Number ______Date______

The respondent has been informed about this survey and understands its purpose and objectives

Signature the respondent ______Signature of the interviewer ______

* Modified consent form, form the Guidelines for measuring National TB/HIV prevalence in population based surveys, 2005. WHO/UNAIDS.

212

Physician's Knowledge Questionnaire

A. Baseline Data

1. Name ______2. Gender [1] Male [2] Female 3. Age_____Years 4. BHU______5.District ______6. Contact No ______7.Date of Entry ______B. Interview

I. Demographic Information 1. Medical School attended [1] Government [2] Private [3] Foreign

2. Qualification [1] MBBS [2] MD [3] MBBS + Post Graduation [4] MD + Post Graduation

3. Year of Graduation ______

4. Do you have any Experience in Pulmonology? [1] Yes [2] No

5. What Experience and for how long? ______

6. Duration of Service ______years

7. Did you receive any training on TB? [1] Yes [2] No

8. Do you need Training on TB? [1] Yes [2] No

213

II Knowledge and Understanding of Tuberculosis (TB)

1. What Germs Causes TB? ______

What is the Most Common Symptom of TB? ______

2. How you will diagnose TB? ______

3. Do you think TB is Communicable Disease? [1] Yes [2] No [3] Don't Know

4. Do you know how TB Spread? ______

5. What is the minimum duration of TB Treatment? ______

6. How you will plan TB treatment? [1] ______[2] ______

7. How you will treat a pregnant woman with TB? ______

8. How you will treat a 15 years (25 Kg) child with TB? ______

9. Do you know the recommended doses of TB drugs? [1] Rifampacin ______[2] Isoniazid ______[3] Ethambutal ______[4] Pyrazinamide _____ [5] Streptomycin ______214

10. What is TB DOTS? ______

11. How you will do follow up of TB patients? ______

12. Which TB drugs are contraindicated in Hepatitis? ______

13. What is MDR-TB? ______

14. How you will treat MDR-TB? ______

215

14.5 Conference Presentations

216

217

Abstract

Do you know women are at a higher risk of Tuberculosis (TB) than male in Khyber Pakhtunkhwa (KPK), Pakistan?

M Aziz, D Baxter, A Esmail, 1Primary Care Research, School of Medicine, University of Manchester UK

Introduction: A descriptive epidemiological study of Tuberculosis in the province of KPK, Pakistan between 2002 and 2010. Research provides important data regarding the diseases burden and to confirm that women are at a higher risk for TB in the region.

Methods: We retrospectively collected data for all registered TB patients between 2002 and 2010 in KPK using National TB Control Programme (NTP) database. Access to database was provided by the NTP Information technology (IT) department with valid user name and password. The data were analysed through Microsoft Excel 2007 and tabulated in aggregated style to show the distribution of TB by age, gender, type of disease, geographical location and treatment outcome. Central tendency and variability parameters included mean, median and standard deviation obtained using SPSS 17. Qualitative data reported in absolute rate percentages.

Results: Total of 69,387 new TB patients were reported, with 57% female and 42% male. Majority of cases registered in Peshawar District 11836 (56.41% female, 43.59% Male). The lowest no of cases reported from Hungu District 646 (59.24% Female, 40.87% Male). Kohistan District showed the greatest gender difference with female constituting 78.80% and 22% male. Another five districts Battagram, Chitral, Buner, Upper Dir and Swabi reported that women accounted for more than 60% of the total TB cases.

Discussion: We confirmed that women of Khyber Pakhtunkhwa are at higher risk of TB than male. Further research is in progress to explore the causes of this higher incidence as a PhD project in the University of Manchester, UK.

Funding: We thanks to The University of Manchester United Kingdom, Islamic Development Bank Jeddah Saudi Arabia and National Tuberculosis Control programme Khyber Pakhtunkhwa for funding and providing technical support for the study.

Competing Interests: It has been declared that no competing interests exist by the authors for this study.

218

219

220

Abstract Tuberculosis in Adult Women: A systematic Review of Gender Differences in Tuberculosis Notification in the province of Khyber Pakhtunkhwa, Pakistan. 1M Aziz, 2M Ali, 1A Esmail, 1Primary Care Research, School of Medicine, University of Manchester UK 2National Tuberculosis Control Programme Khyber Pakhtunkhwa, Pakistan.

Background: Mycobacterium Tuberculosis (TB) is a serious public health threat globally and particularly in developing countries. Among the 22 high burden countries, Pakistan ranks 5th with a diagnosis of about 420,000 new cases every year. Globally, TB case notification is high in male as compared to female but in Khyber Pakhtunkhwa (PK) female patients are at a higher risk of TB than male. Similar data reported from Afghanistan, where 68.50% new sputum positive patients were female. Aims and Objectives: The purpose of this systematic review of published studies was to explore the causes of gender differences in the notification of Tuberculosis in KPK. The review formulates a theoretical base for my research identifying key areas for data collection (Social and Epidemiological Risk factors, Patients knowledge and Physicians Knowledge).

Methodology: A relevant literature was identified through covering key theoretical papers and review articles on risk factors for Tuberculosis and gender issues. Databases accessed were MEDLINE, EMBASE and GLOBAL HEALTH and searches were conducted focusing on literature published between January 2000 and February 2015. The websites of specific journals, including the Journal of Infectious Diseases and American Journal of Public Health (AJPH) were used to locate references cited in articles (mostly review articles). Additionally, the websites of the World Health Organisation (WHO), the National Tuberculosis Control Programme (NTCP) and the Department of Health, Pakistan, were used to find related reports. Data were extracted from selected 85 research articles for tabulation, comparison and synthesis.

Results: A total of 85 research articles were selected from 4,630 screen citations. Evidence extracted from these articles was used for metasynthesis of hypothetical base. Different main themes were noted reviewing the literature; age and gender; educational status of TB patients; economic status; nutritional status; overcrowding and household status; fuel consumption; rural and urban residence; knowledge of patients regarding TB and knowledge of general practitioners about diagnosis, treatment and fallow up. Our systematic analysis of the literature explained how socioeconomic issues interrelate to influence TB management. Results of our review are limited due to personal selection and quality of articles.

Conclusions: The results of our literature review showed that very limited literature available on gender differences in TB. Gender and TB is a complex and neglected area of research in high burden countries. In spite of higher female notification of TB in KPK since 2000, no study has been carried out to address this issue. The results of this study will help National TB Control Programme to initiate a gender specific research in KPK.

221

14.6 University Positions for Students

222

223

224

PGR Student Representaives University of Manchester (School of Medicine)

225

14.7 National TB Control Programme Data Collection and Reporting Instruments

226

227

228

229

230

231

232

233

234

14.8 PRISMA Checklist

235

Reported on Section/topic # Checklist item page # TITLE Title 1 Tuberculosis in Adult Women: A systematic Review of Gender Differences in Tuberculosis 53 Notification in the province of Khyber Pakhtunkhwa, Pakistan.

ABSTRACT Structured 2 The results of our literature review showed that very limited literature available on gender 59 summary differences in MTb. Gender and MTb is a complex and neglected area of research in high burden countries. In spite of higher female notification of MTb in KPK since 2000, no study has been carried out to address this issue. The results of this study will help National TB Control Program to initiate a gender specific research in KPK

INTRODUCTION Rationale 3 Mycobacterium Tuberculosis (MTb) is a serious public health threat globally and particularly in 53 developing countries. Among the 22 high burden countries, Pakistan ranks 5th with a diagnosis of about 420,000 new cases every year. Globally, MTb case notification is high in male as compared to female but in Khyber Pakhtunkhwa (KPK) female TB notification is higher as compared to male. Similar data reported from Afghanistan, where 68.50% new sputum positive patients were female. Objectives 4 The purpose of this systematic review of published studies was to explore the causes of gender 53 differences in the notification of Tuberculosis in KPK. The review formulates a theoretical base for my research identifying key areas for data collection (Social and Epidemiological Risk factors, Patients knowledge and Physicians Knowledge), using PRISMA-P METHODS Protocol and 5 PRISMA-P 59 registration

236

Eligibility 6 Empirical research or a systematic review relevant to the epidemiology and risk factors for TB. 58 criteria Articles published in English Knowledge, Attitude, Awareness and Treatment seeking behaviour in TB patients-related papers. Assessment of learning, approach and management of TB amongst family physicians-related papers. Information 7 Databases: MEDLINE, EMBASE and GLOBAL HEALTH Journals: The Journal of Infectious 54 sources Diseases and American Journal of Public Health (AJPH) Websites: Health Organisation (WHO), the National Tuberculosis Control Programme (NTCP) and the Department of Health, Pakistan.

Search 8 The search strategy sought material evaluating epidemiology and gender differences, risk factors, 55-58 patient knowledge and general practitioners knowledge of TB in Pakistan and Khyber Pakhtunkhwa. Key words “Tuberculosis or TB”, “risk factors”, “adult female or females”, “Pakistan” and “Khyber Pakhtunkhwa or NWFP” were used to search at the national level. I also used key words ”China”, “India”, “Iran”, “Afghanistan” for available studies to compare the situation of the disease at the regional level as well. Study selection 9 The systematic review identified 4,630 screen citations using multiple approach search strategy. 59 Of the total, 2530 studies yielded by MEDLINE search, EMBASE search yielded 1155 and GOLBAL HEALTH 730 articles only. Additionally, from the websites of WHO, NTCP, MOH and online journals (JID, AJPH) further 250 articles were identified. The preliminary search reviewed, 3067 articles were excluded as not relevant to the research area. Further reviewing the findings by titles and abstracts, 749 more research articles were removed. After reading full text articles, 154 studies were considered eligible on the basis of our inclusion criteria. Finally, after removing the duplicate articles 85 studies were remained for literature review Data collection 10 A piloted questionnaire was made on the basis of variables reported on literature and data 207-14 process collected for most important variables such as age, gender, educational and economic status etc. The piloted questionnaire was updated after literature review as advised by the supervisory team. 237

Data items 11 List of variables 72-91

Summary 12 Summary tabulated as tables 12,13,14,15 72-80 measures RESULTS Results 13 A total of 85 studies were selected for details please see a flow diagram. Summary tabulated as 59 tables 12,13,15 and 15 61-69 DISCUSSION Summary of 14 The literature on patient's knowledge about Tuberculosis is very limited. In Khyber Pakhtunkhwa, 72-92 evidence I could not identify any research on the knowledge of TB patients. However, there were several studies carried out on patient's knowledge in Pakistan. It is evident from the results of studies included in the literature review that patient's knowledge regarding TB is very limited, especially in rural communities. Since nearly 75 % of KPK population live in rural areas, it is a reasonable assumption that patients from rural districts of KPK have poor knowledge of TB and should be investigated for the association between knowledge and higher notification of TB in females Limitations 15 Literature regarding gender and tuberculosis is very limited especially in KPK 76-77 Conclusions 16 Data from National TB control program and research done in KPK suggested that female 77 notification of TB is higher than male and further research should be carried out to explore the reasons behind this variation. FUNDING Funding 17 Islamic Development Bank Jeddah Saudi Arabia (IDB)

238